The Labour Market Impact of Immigration: Quasi-Experimental … · 2011-09-15 · Denmark and...
Transcript of The Labour Market Impact of Immigration: Quasi-Experimental … · 2011-09-15 · Denmark and...
Centre for Research and Analysis of Migration Department of Economics, University College London Drayton House, 30 Gordon Street, London WC1H 0AX
Discussion Paper Series
CDP No 12/06
The Labour Market Impact of Immigration: Quasi-Experimental Evidence
Albrecht Glitz
Centre for Research and Analysis of Migration Department of Economics, Drayton House, 30 Gordon Street, London WC1H 0AX
Telephone Number: +44 (0)20 7679 5888 Facsimile Number: +44 (0)20 7916 2775
CReAM Discussion Paper No 12/06
The Labour Market Impact of Immigration:
Quasi-Experimental Evidence
Albrecht Glitz*
* Universitat Pompeu Fabra
Non-Technical Abstract
With the fall of the Berlin Wall, ethnic Germans living in the former Soviet Union and the Warsaw Pact countries were given the chance to migrate to Germany. Within 15 years, 2.8 million individuals moved. Upon arrival, these immigrants were exogenously allocated to different regions by the administration in order to ensure an even distribution across the country. Their inflows can therefore be seen as a natural experiment of immigration, avoiding the typical endogeneity problem of immigrant inflows with regard to local labour market conditions. I analyse the effect of these exogenous inflows on relative skill-specific employment and wage rates of the resident population in different geographical areas between 1996 and 2001. The variation I exploit in the empirical estimations arises primarily from differences in the initial skill composition across regions. Skill groups are defined either based on occupations or educational attainment. For both skill definitions, my results indicate a displacement effect of around 4 unemployed resident workers for every 10 immigrants that find a job. I do not find evidence of any detrimental effect on relative wages. Keywords: Immigration, Labour Market Impact, Skill Groups, Germany JEL Codes: J21, J31, J61
The Labour Market Impact of Immigration:Quasi-Experimental Evidence
Albrecht Glitz∗Universitat Pompeu Fabra
January 2008
Abstract
With the fall of the Berlin Wall, ethnic Germans living in the former Soviet Unionand the Warsaw Pact countries were given the chance to migrate to Germany. Within15 years, 2.8 million individuals moved. Upon arrival, these immigrants were ex-ogenously allocated to different regions by the administration in order to ensure aneven distribution across the country. Their inflows can therefore be seen as a naturalexperiment of immigration, avoiding the typical endogeneity problem of immigrantinflows with regard to local labour market conditions. I analyse the effect of these ex-ogenous inflows on relative skill-specific employment and wage rates of the residentpopulation in different geographical areas between 1996 and 2001. The variationI exploit in the empirical estimations arises primarily from differences in the initialskill composition across regions. Skill groups are defined either based on occupationsor educational attainment. For both skill definitions, my results indicate a displace-ment effect of around 4 unemployed resident workers for every 10 immigrants thatfind a job. I do not find evidence of any detrimental effect on relative wages.
Keywords: Immigration, Labour Market Impact, Skill Groups, Germany
JEL Codes: J21, J31, J61
∗Department of Economics and Business, Ramon Trias Fargas 25-27, 08005 Barcelona, Spain, e-mail:[email protected]. This paper was written while working at University College London and the Centrefor Research and Analysis of Migration (CReAM). I am grateful to Christian Dustmann, David Card, Ken-neth Chay, Emilia Del Bono, Ian Preston, Imran Rasul, Regina Riphahn and Matti Sarvimaki for helpfulcomments and suggestions, and to Stefan Bender for invaluable support with the data. I have benefitedfrom many useful comments of conference participants at ESPE 2004, EEA 2005, the COST A23 meet-ing 2005, the CReAM/TARGET conference 2006, EEA 2006, EALE 2006 and participants of the LaborLunch Seminar at Berkeley. Parts of this paper were written while visiting the Department of Economicsat Berkeley, which I thank for the hospitality. I also thank the ESRC for funding the project (award No.RES-000-23-0332) and the Barcelona Economics Program of CREA for its support.
1 Introduction
The impact immigration has on the labour market outcomes of the resident population is
a central issue in the public debate on immigration policies. In most European countries
it has been widely discussed in recent years in connection with the eastern enlargement
of the European Union and, in particular, the potential introduction of transitional mea-
sures to restrict labour migration from the new member states. There is a widespread
concern that immigrants exert downward pressure on wages and reduce job opportunities
for resident workers. Since the 1990s, numerous studies have tried to empirically assess
the labour market effects of immigration for a number of countries, sometimes with con-
flicting results and using a variety of methodological approaches.1 The most common
approach in the literature is the spatial correlation approach, in which a measure of the
employment or wage rate of resident workers in a given area is regressed on the relative
quantity of immigrants in that same area and appropriate controls.2 One of the main diffi-
culties of this strategy arises from the immigrants’ potentially endogenous choice of place
of residence. Immigrants tend to move to those areas that offer the best current labour
market opportunities, which typically leads to an underestimation of the true effect they
have on the labour market outcomes of the resident population. To address this endo-
geneity problem, some studies have used instrumental variables that are based on past
immigrant concentrations, exploiting the fact that these are good predictors of contempo-
rary immigrant inflows while assuming that they are uncorrelated with current unobserved
labour demand shocks.
In this paper, I follow an alternative approach by taking advantage of a natural experiment
in Germany in which a particular group of immigrants was exogenously allocated to spe-
1See Friedberg and Hunt (1995), Gaston and Nelson (2002) or Dustmann and Glitz (2005) for compre-hensive surveys of the literature.
2Examples include Altonji and Card (1991), LaLonde and Topel (1991), Butcher and Card (1991), andCard (2001) for the U.S., Winter-Ebmer and Zweimuller (1996, 1999) for Austria, Hunt (1992) for France,Pischke and Velling (1997) for Germany, Carrington and de Lima (1996) for Portugal, Dustmann et al.(2005) for the UK, and Hartog and Zorlu (2005) for the Netherlands, the UK and Norway.
1
cific regions upon arrival by government authorities. The prime objective of the allocation
policy was to ensure an even distribution of these immigrants across the country. Since,
to an overwhelming extent, the actual allocation decision was based on the proximity of
family members and sanctions in case of non-compliance were substantial, the possibility
of self-selection into booming labour markets was severely restricted for this group of
immigrants, allowing us to view their settlement as exogenous to local labour market con-
ditions and providing a unique opportunity to study its effect on the resident population.
Only in few instances is it feasible to view immigration as a natural experiment in which
the immigrant inflows into a particular region are not driven by local labour market condi-
tions. The only example in the literature that uses such an experiment to identify the labour
market impact of immigration on the resident population is the Mariel boatlift analysed
by Card (1990).3 The main conceptual difference between that study and my analysis is
that Card examines a large exogenous inflow into a single local labour market, the city of
Miami, whereas this analysis uses exogenous but relatively homogenous inflows into all
regions in Germany. As I will show, in this case the main source of variation stems from
differences in the skill composition of the resident labour force across regions. Edin et al.
(2003), Piil Damm (2006) and Gould et al. (2004) are further studies that are related to my
analysis insofar as they use spatial dispersal policies for refugee immigrants in Sweden,
Denmark and Israel, respectively, as a source of exogenous initial regional allocations of
immigrants. Rather than looking at the labour market impact of these inflows on the res-
ident population, the aim of the former two studies is to assess how living in an ethnic
enclave affects immigrants’ own labour market outcomes whereas the latter investigates
the effect of school quality on the high school performance of immigrant children.
In this paper, I set up a model in which immigration affects the relative supplies of differ-
3There are a number of studies, however, in which the immigrant inflow to a country as a whole - ratherthan to particular regions within the country - can be seen as a natural experiment, for instance the inflow ofrepatriates from Algeria to France analysed by Hunt (1992) or the mass migration of Russian immigrants toIsrael studied by Friedberg (2001).
2
ent skill groups in a locality. I then estimate how changes in these relative supplies affect
the employment/labour force rate and wages of the resident population, first by OLS and
then using the exogenous immigrant inflows to instrument the potentially endogenous
changes in relative skill shares in a locality. I define skill groups in two alternative ways
based on either occupations or educational attainment and distinguish between the effect
on native Germans and foreign nationals. To investigate whether out-migration of the res-
ident population in response to the immigrant inflows potentially dissipates their labour
market impact across the economy, I regress overall and skill-specific local population
growth rates on immigrant inflow rates. The results from these regressions also allow an
assessment of whether there is any positive association between immigrant inflows and
the growth rates of the resident population, which would cast doubt on the exogeneity
of the allocation decisions with regard to local demand conditions. Finally, I ascertain
whether the initial skill composition in a locality, which turns out to be the main source of
variation in my estimations, has an independent effect on future changes in labour market
outcomes that could be driving the results.
The particular group of immigrants at the centre of this study are so called “ethnic Ger-
man immigrants”. These are individuals who were living in large numbers in Central and
Eastern Europe and the former Soviet Union and who were particularly affected by the
divisive ideological developments in the aftermath of World War II. Only as a result of the
political changes in the former Eastern Bloc towards the end of the 1980s did this group
gain the opportunity to immigrate to Germany, which, after 40 years of isolation, was
eagerly seized. Between 1987 and 2001 more than 2.8 million ethnic German immigrants
moved to Germany, increasing its population by 3.5%. Based on Germany’s principle of
nationality by descent, this particular group of immigrants as well as their descendants
are regarded as German by the constitution and granted German citizenship in the event
of immigration. I collected annual county-specific inflows of this group of immigrants di-
rectly from each of the sixteen federal admission centres and combine these figures with
3
detailed information on local labour markets that I obtained from social security based
longitudinal data. The analysis focuses on West Germany, excluding Berlin, and covers
the period 1996 to 2001, during which the allocation policy was in effect.
The empirical results point towards the existence of unobserved local demand shocks that
are correlated with changes in relative skill shares and lead to upward biased estimates
of the labour market impact of immigration from simple OLS regressions. Using the eth-
nic German immigrant inflows to instrument the endogenous changes in the relative skill
shares leads to substantially larger negative effects on the employment/labour force rate.
The estimates imply that for every 10 immigrant workers finding employment, about 4
resident workers lose their jobs. Since all regressions are based on annual variation, this
displacement effect has to be interpreted as a short-run effect. The increase in magnitude
of the estimates by a factor of 3 to 7 when moving from OLS to IV is comparable with
the results Card (2001) found in a similar study for the U.S., in which the instrument,
however, was based on past immigrant settlement patterns. The fact that I find a negative
effect on the employment/labour force rate of the resident population stands in contrast to
a number of earlier studies for Germany, for instance to Pischke and Velling (1997) and
Bonin (2005), who do not find such effects. My results do not show evidence of detri-
mental effects on relative wages of the local population. Finally, there is no indication
that the obtained results are underestimates of the immigrant labour market impact due to
compensatory outflows of the resident population or that they are driven by an indepen-
dent effect of initial relative skill shares on future labour market outcomes.
The remainder of this paper is organised as follows. In the next section, I will provide
some background information on ethnic German immigration since World War II and the
institutional setting in which it took place. In Section 3, I explain the underlying theo-
retical model and identification strategy of my analysis. I then describe the data sources
in Section 4 and provide some descriptive evidence in Section 5. Finally, I present and
4
discuss the estimation results in Section 6. Section 7 concludes.
2 The German Migration Experience - Some Facts
2.1 Historical Background
Figure 1: Ethnic German immigrant inflows by country of origin, 1950 to 2001
010
0000
2000
0030
0000
4000
00
1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000Year
Poland Former Soviet UnionRomania Former CSSR, Hungary and other countries
Source: Bundesverwaltungsamt
To understand the origin of ethnic German immigrants we have to consider their his-
torical background. During the terror regime of the National Socialists in Germany, a
large number of German citizens fled the country or were forcibly resettled to the east-
ern occupied territories. After the end of World War II and the ensuing repartitions and
forced resettlements across Europe, about 15 million German citizens became refugees
or expellees, most of whom moved back to Germany in the immediate post-war years.
According to Salt and Clout (1976) some 7.8 million of these refugees had settled in West
Germany and 3.5 million in East Germany by 1950. However, many German citizens and
5
their descendants continued to live outside post-war Germany. Their inflows gradually
ebbed away as Eastern European countries became increasingly isolated. After the ini-
tial post-war displacements, immigration of ethnic Germans, then called Aussiedler, took
place on the basis of bilateral agreements between Germany and the corresponding source
countries. However, after the construction of the Berlin Wall in 1961 and the worsening
of the East-West relations, these flows were severely limited. Between 1950 and 1987,
the number of ethnic Germans who came to West Germany added up to 1.4 million, of
which 848,000 had come from Poland, 206,000 from Romania, and 110,000 from the for-
mer Soviet Union.4 In 1988, with the end of the cold war looming, travel restrictions in
Central and Eastern Europe were lifted. This caused an immediate resurgence of ethnic
German migrations. In 1990 alone some 397,000 individuals, mainly from the former
Soviet Union (37%), Poland (34%) and Romania (28%), arrived in Germany (see Figure
1). Faced with these enormous movements, the government limited their inflow in sub-
sequent years at a level of around 225,000 per year. This quota was met until 1995 after
which the annual inflows gradually decreased. From 1993 onwards more than 90% of
the ethnic German immigrants originated from territories of the former Soviet Union. It
is important to emphasise that the ethnic German immigrant population I analyse in this
study does not include Germans who used to live in East Germany and who moved to
West Germany after unification in 1990. This group had complete freedom of movement
within Germany from the day of unification.
2.2 Institutional Framework
All ethnic German immigrants who want to come to Germany have to apply for a visa
at the German embassy in their country of origin and prove their German origin in terms
of descent, language, education and culture. Once applications are accepted and a visa is
granted, which takes around one year, all arriving immigrants have to pass through a cen-
tral admission centre where they are initially registered. In case they do not have a job or
4Source: Bundesverwaltungsamt, Jahrestatistik Aussiedler 2003.
6
other source of income that guarantees their livelihood, which applies to the vast majority
of immigrants at the time of arrival, they are then allocated to one of the sixteen federal
states according to pre-specified state quotas.5 Within each state, they are subsequently
further allocated to particular counties, using a state-specific allocation key as guidance
which, with two exceptions, is fixed over time and based on the relative population share
of each county.6 By far the most important factor determining the final destination of the
ethnic German immigrants is the proximity of family members or relatives. The responsi-
ble authority at the Ministry of the Interior estimates that this has been the decisive factor
in the allocation decision in approximately 90% of all cases. Additional factors are the
presence of health and care facilities and the infrastructure for single parents. Crucially
for this study, the skill level of the immigrants did not play any substantial role in the
allocation process.
The legal basis for this system is the “Assigned Place of Residence Act” (Wohnortzuweisungs-
gesetz), which was introduced in 1989 in response to the large inflows experienced at the
time. These inflows tended to be concentrated towards a few specific regions where they
caused considerable shortages in available housing space while in other, particularly rural
areas, facilities remained empty.7 The intention of the law was to ensure a more even dis-
tribution of ethnic German immigrants across Germany and avoid a capacity overload of
local communes, who are responsible for the initial care of the immigrants. However, in
practice, the introduction of this law turned out to be ineffective because the entitlements
5According to the so-called Konigsteiner Distribution Key, the quotas since 1993 have been: Baden-Wurttemberg 12.3%, Bavaria 14.4%, Berlin 2.7%, Brandenburg 3.5%, Bremen 0.9%, Hamburg 2.1%, Hesse7.2%, Mecklenburg-Pomerania 2.6%, Lower Saxony 9.2%, North Rhine-Westphalia 21.8%, RhinelandPalatinate 4.7%, Saarland 1.4%, Saxony 6.5%, Saxony-Anhalt 3.9%, Schleswig-Holstein 3.3%, andThuringia 3.5%.
6The exceptions are Lower Saxony where the quotas are annually adjusted for changes in each county’spopulation, and North Rhine-Westphalia where quotas are based on both population and geographical areaand annually adjusted to population changes.
7The problem of housing space was particularly pronounced in the late 1980s and early 1990s whenannual inflows of ethnic German immigrants were largest. By the mid 1990s, however, sufficient capacitiesin social housing and hostels had been established and were even partly shut down again due to the smallerannual inflows. Therefore I do not expect that housing availability, which may depend directly on the stateof the local economy, would have affected the number of immigrants allocated to a region and in that wayintroduced endogeneity into the allocation process.
7
to considerable statutory provisions such as financial social assistance, free vocational
training courses, and language classes were not affected should the ethnic German im-
migrant choose to settle in a region different from the one allocated upon arrival. As
a consequence, unregulated internal migration of ethnic Germans led to the creation of
a few enclaves, in some of which their concentration reached up to 20% of the overall
population (Klose, 1996). In response to these developments, the Assigned Place of Res-
idence Act was substantially modified on 1 March 1996. As a key feature of the new
law, ethnic German immigrants would now lose all their statutory entitlements in case of
non-compliance with the allocation decision. Due to the federal structure of Germany it
was subject to each of its states to adopt and implement the new legislation. Apart from
Bavaria and Rhineland-Palatinate, all West German states chose to do so, most of them
with effect from 1 March 1996. Only Lower Saxony and Hesse adopted the law at a later
point, the former in April 1997 and the latter in January 2002. For an overview see Table
B-1 in Appendix B. The perception at both the Ministry of the Interior as well as the
Association of German Cities and Towns is that the new provisions and sanctions have
been successful and ensured a high compliance with the initial allocation decision.8
The regional allocation of the ethnic German immigrants becomes void if they can verify
that they have sufficient housing space as well as a permanent job from which they can
make a living, at the latest, however, three years after initial registration. This suggests
that after arrival in the allocated place of residence there is some scope for endogenous
self-selection through onward migration. However, it is likely that immigrants will pre-
dominantly search for job opportunities in the vicinity of their places of residence. In fact,
the difficulties of searching for a job in a different locality arising from the legal provi-
sions of the Assigned Place of Residence Act were acknowledged by the legislator and
led to a further amendment of the law on 1 July 2000 that explicitly allowed for temporary
8This is corroborated in the commentarial statement of a related judgment by the Federal ConstitutionalCourt in a case in which an ethnic German immigrant took legal action without avail against the restrictionof her freedom of movement (BVerfG, 1 BvR 1266/00 vom 17.3.2004, Absatz-Nr. 1 - 56).
8
residence in alternative localities for the purpose of job search activities without loss of
entitlements as long as it did not exceed 30 days.9
To sum up, through the introduction of the new legislation in 1996 the authorities imple-
mented a system to allocate a particular group of immigrants exogenously with regard
to their skill levels across different regions while at the same time providing for the nec-
essary sanctions to ensure compliance with these allocation decisions. This framework
can therefore be regarded as a natural experiment of immigration in which inflows are
exogenous to local labour demand conditions.
3 Theory
3.1 Empirical Model
The empirical analysis in this paper is based on a model in which immigration impacts
local labour markets by changing the relative supplies of different skill groups (compare
Card, 2001). Assuming that in each labour market a competitive industry produces a
single output good using a CES-type aggregate of skill-specific labour inputs as well as
capital, relative wages and, by substituting into a labour supply function, relative employ-
ment rates will only depend on the relative supply of each skill group.10 The equations
for the effect on the employment/labour force and wage rates are then given by
∆ log(N jrt/Pjrt) = v′jt + v′rt +β1∆ log f jrt +∆v jrt (1)
9I do not explicitly take this change in regulations into account in the analysis since it was only validfor the last six months of the six-year period I cover and did not affect the initial allocation to a particularregion.
10The key assumptions underlying this model are that capital and labour are separable in the local pro-duction function, that the elasticities of substitution across all skill groups are identical, that natives andimmigrants are perfect substitutes within skill groups, and that the per-capita labour supply functions forthe different skill groups have the same elasticity.
9
∆ logw jrt = u′jt +u′rt +β2∆ log f jrt +∆u jrt , (2)
where ∆ log f jrt = log(Pjrt/Prt)− log(Pjrt−1/Prt−1) denotes the percentage change in the
fraction of the overall labour force in labour market r that falls into skill group j, and v′jt ,
u′jt , v′rt , and u′rt are interactions of skill group and year fixed effects and region and year
fixed effects, respectively. ∆v jrt and ∆u jrt are unobserved error components that capture
skill-, region- and year-specific productivity and demand shocks. For a detailed derivation
of these equations see Appendix C.
As opposed to Card’s study, which only uses one cross-section and thus estimates in lev-
els, I am able to control for skill region specific fixed effects (which I difference out) and
use variation in local skill shares over time to identify β1 and β2. This could potentially
be important since otherwise any instrumental variable that is based on past labour market
characteristics will be invalid if these characteristics are themselves correlated with unob-
served skill region specific fixed effects.11
Equations 1 and 2 relate changes in the local employment and wage rates to changes in
the relative factor shares in a locality. Any skill-specific local productivity and demand
shocks in a given year are captured in the error component. If these shocks raise employ-
ment and wage rates in a particular skill group and at the same time attract more workers
into that group, this will induce a positive correlation between the error terms ∆v jrt and
∆u jrt in Equations 1 and 2 and the change in the relative skill share ∆ log f jrt . In this case,
OLS estimates of β1 and β2 will be upward biased.
To address this problem, I take advantage of the exogenous allocation of ethnic German
11If, as for the U.S. and Germany, immigration has historically been unskilled, then it is likely thatany (un)skilled region fixed effect is correlated with the overall number of immigrants living in a locality:unskilled immigrants would have tended to move to those areas that are particularly attractive given theirskill level. In a cross sectional analysis skill region fixed effects cannot explicitly be controlled for and arepart of the unobserved error component. An instrument that is based on past immigrant concentrations willthen be correlated with this error component, rendering it invalid.
10
immigrants to Germany’s counties between 1996 and 2001. Specifically, I assume that
their inflows are uncorrelated with any skill-specific productivity and demand shocks and
can therefore serve as an instrument for the change in the relative factor shares ∆ log f jrt .
I will provide evidence for the validity of this assumption in Section 5.4.
I construct my instrument, the skill-specific ethnic German inflow rate, by multiplying the
overall inflow ∆Irt into a particular locality with the nationwide fraction of ethnic German
immigrants in each skill group where I distinguish skill groups either by educational at-
tainment or by occupation. Let θ jt denote this fraction and let ωt denote the fraction of
ethnic German immigrants that arrive in year t and are aged between 15 and 64. Since
individual skills and age did not play a role in the allocation of ethnic Germans to local
labour markets, one can expect the skill and age composition of the arriving ethnic Ger-
man immigrants in each locality to be the same.12 The predicted skill-specific inflow rate
of working age immigrants into labour market r in year t that I use as an instrument for
the change in the relative factor share is then given by
SPjrt =θ jtωt∆Irt
Pjrt−2,
where SPjrt stands for the skill-specific supply-push component of ethnic German immi-
grant inflow ∆Irt , and Pjrt−2 is the overall labour force in skill group j in t−2. I use a lag
of two years in the denominator in order to avoid any correlation with the skill-specific
error terms ∆v jrt and ∆u jrt in Equations 1 and 2.13
12In the presence of a correlation in skills between immigrants and their family contacts already living inGermany, this assumption may not hold. However, since these families have typically been split up a longtime ago and passed through significantly different educational systems, the correlation in skills is likelyto be small. If the assumption of identical skill compositions of arriving ethnic Germans were invalid, thiswould be reflected in a weak first stage of the instrumental variable estimations.
13Using the skill-specific labour force of the previous year instead would increase the first stage correla-tion of the instrument with the endogenous variable ∆ log f jrt but, in the presence of unobserved productivityand demand shocks, introduce a positive correlation of the instrument with the first differenced error terms∆v jrt and ∆u jrt which would render the instrument invalid. For the skill-specific labour force of the previousyear to be valid for the construction of the instrument would require that the employment/labour force rateevolves as a random walk, a requirement unlikely to hold for Germany (see Pischke and Velling, 1997, fora discussion of this issue).
11
Based on my data, the skill-specific labour force in a locality consists of all employed
individuals plus all individuals receiving official unemployment compensation, either un-
employment benefits (Arbeitslosengeld) or unemployment assistance (Arbeitslosenhilfe).
During the period covered by this analysis, unemployed individuals receive unemploy-
ment benefits for the first 6 to 32 months dependent on the duration of their previous
employment. Subsequently, they receive unemployment assistance which is means-tested
and, in principle, indefinite. The data therefore provides a fairly good approximation of
the actual labour force, in particular for men which are less likely to lose or quit their
job without receiving some sort of unemployment compensation thereafter. A peculiarity
arising from these data with respect to the empirical model, however, is that year to year
changes in the local skill shares are driven by new individuals becoming employed in a
given skill group. This is because in order to qualify for official unemployment compen-
sation individuals first have to work for at least 12 months prior to becoming unemployed,
so that new entrants into the labour force always “enter” my data set as employed individ-
uals.14 This has an important implication for the interpretation of the coefficients β1 and
β2. These now measure how changes in the relative skill shares in a locality induced by
additionally employed individuals affect average labour market outcomes. In the case of
the employment/labour force rate, β1 hence measures the direct displacement effect, that
is, how many workers lose their job for every additional worker finding a job.
3.2 Source of Variation
An important issue in the context of this study is that, by design, the exogenous alloca-
tion of ethnic German immigrants over the entire German labour market ensures that the
variation in the overall regional inflow rates is small. In fact, if the overall number of
14In the data, the recorded locality for an unemployed individual always corresponds to the locality of theprevious employment spell. The only way the relative skill share in a locality can then change by additionsto the number of unemployed from one year to the next is when an already eligible worker moves into a jobin a new locality but then becomes unemployed before the cut-off date at which I calculate the relative skillshares.
12
immigrants allocated to each county was strictly proportional to the resident population,
there would be no variation in the overall ethnic German immigrant inflow rate and simply
regressing local labour market outcomes on the overall inflow rate, as done in many im-
pact analyses (for instance Altonji and Card, 1991 or Pischke and Velling, 1997), would
have been impossible. Moreover, if the allocation decision is based, as in the present case,
to an overwhelming extent on family ties, the skill distribution of the newly arriving ethnic
German immigrants is also going to be homogeneous across different regions. However,
even with the same inflow rate and skill composition of the arriving immigrants in each
region, the effect on the labour market outcomes of the resident population of a particular
skill group will still differ dependent on the existing pre-migration skill distribution in
each region. In particular, the percentage change in local skill share f jrt after an inflow
of immigrants that is homogenous across regions r relative to the resident population,
∆IrtPrt−1
= it , and of which a constant share across regions of v jrt = v jt is of skill j is given
by
%∆ f jrt =f jrt−1 + v jt itf jrt−1(1+ it)
−1, (3)
where, for simplicity, I assume that there is no growth in the local population for other
reasons than immigration. The first derivative of this term with respect to the initial skill
share f jrt−1 is then given by
− v jt itf 2
jrt−1(1+ it)< 0,
so the larger the initial skill share, the smaller will be the percentage change in the relative
skill supply induced by the skill-homogenous inflow of immigrants.
Differences in the skill composition before the immigrant inflows occur thus lead to differ-
ences in the relative changes of the skill shares and hence to differences in the responses
of labour market outcomes. The variation I exploit in my estimations therefore arises
13
Figure 2: Source of variation
mainly from variation in the pre-existing skill compositions across different labour mar-
ket regions rather than from a differential composition of the immigrating population.
Figure 2 illustrates this point. Suppose there are two regions, Region A and Region B,
where Region A is a low skill region with 80% of the workforce being low-skilled, 15%
medium-skilled, and 5% high-skilled while Region B is a high skill area with 5% low-
, 15% medium-, and 80% high-skilled. Suppose skill is here measured by educational
attainment. Now suppose there is a 1% inflow into each region of which 43% are low-
skilled, 46% medium-skilled and 10% high-skilled. The values here reflect the corre-
sponding skill shares in our immigrating population. Such an inflow will now lead to
significantly different changes in relative skill shares in Regions A and B. While in Re-
gion A the share of low-skilled workers will decrease by -0.5%, it increases by 7.6% in
Region B. Conversely, the inflow of high-skilled immigrants will lead to a 1% increase
in the share of high-skilled individuals in Region A and a -0.9% reduction of the share in
Region B. Given our model, it is the percentage changes in relative skill shares that are
driving labour market outcomes and which provide the variation we use to identify the
labour market impact of immigration.
14
4 Data Sources
4.1 Data on Ethnic German Immigrants
At the end of every year, the Federal Administration Department in Germany (Bundesver-
waltungsamt) publishes information on the recent cohort of ethnic German immigrants
in their series “Jahresstatistik fur Aussiedler”. These publications contain information
recorded upon the immigrants’ arrival in Germany; specifically on their countries of ori-
gin, age structure, last occupation, last labour force participation status, and religious
affiliation. They also include the absolute numbers allocated to each of Germany’s six-
teen federal states. All the information provided is on the national level, apart from the
age structure and religious affiliation, which are detailed for each state separately. Of par-
ticular importance for this analysis is the information on the last occupation in the country
of origin since it provides a measure of the immigrants’ skill levels that is exogenous to
local demand conditions in Germany. I use this occupational information to calculate the
fraction θ jt of ethnic German immigrants in each occupation group, which I require for
the construction of my instrumental variable.
I augment the aggregate information from the annual publications with data on the re-
gional inflows of ethnic German immigrants. Since there is no information on the country
of birth of an individual in my main data source on local labour market characteristics,
these immigrants are not distinguishable from those Germans who were born in Ger-
many (and to which I will henceforth refer as “native Germans”). Tracking where they
actually settled is therefore not possible from these data. For that reason, I approached
the responsible federal admission centres for each state directly, which due to the decen-
tralised allocation process are separately responsible for recording the actual inflows. I
was able to obtain the relevant information for each county in West Germany’s ten federal
states with the exception of Bavaria, where records were not kept at the required regional
15
level.15 The period I cover is from 1996 to 2001 during which the Assigned Place of
Residence Act was in effect. I focus on West Germany (excluding Berlin) since data on
ethnic German inflows to the territory of what was formerly known as the German Demo-
cratic Republic are very fragmentary. Furthermore, local labour markets in that area have
experienced fundamental changes after German unification in 1990 in their transition to
market economies which are difficult to control for and may contaminate the results of
this study.
4.2 German Microcensus
While the last occupation in the country of origin is reported upon arrival in Germany
and published in the annual reports of the Federal Administration Department, there is no
information on the immigrants’ educational attainment. I use the German Microcensuses
of 1999, 2001, and 2002 to obtain this information. In each Microcensus I am able to
identify ethnic German immigrants as individuals with German citizenship that arrived in
Germany in any particular year between 1996 and 2001.16 For any given year of arrival
there were between 94 and 274 individuals aged 15 to 64 with valid educational informa-
tion. From these observations I calculate the fraction θ jt of ethnic German immigrants
in each education group, which again is used for the construction of my instrumental
variable in the regressions based on education groups. Since I am interested in the immi-
grants’ educational level upon arrival, I use the available information closest to the actual
year of arrival. The skill shares for 1996, 1997 and 1998 are therefore taken from the 1999
Microcensus, the shares for 1999 and 2000 from the 2001 Microcensus, and the shares
15The other nine federal states or Lander in West Germany are Schleswig-Holstein, Hamburg, LowerSaxony, Bremen, North Rhine-Westphalia, Hesse, Rhineland-Palatinate, Baden-Wurttemberg and Saarland.
16Unfortunately, there is no information in the Microcensuses on the country of origin so that some ofthe individuals I identify as ethnic Germans could in fact be German citizens immigrating from other, forinstance Western European or North American countries. In an alternative data set, the European SocialSurvey 2003, which does include the necessary information, I am able to identify 33 individuals withGerman citizenship who were not born in Germany and who moved to Germany between 1993 and 2003.All 33 of these ethnic German immigrants came from typical source countries of Aussiedlers, mostly fromKazakhstan (14) and Russia (13). Although the sample is small, it indicates that the share of immigratingethnic Germans from other regions is likely to be small.
16
for 2001 from the 2002 Microcensus.17
4.3 IAB Employment Subsample
I obtain data on the labour market outcomes of the resident population from the Em-
ployment Subsample 1975-2001 which is made available by the Institute for Employment
Research (IAB). This administrative data set comprises a 2% subsample of all dependent
employees subject to social security contributions in Germany. It includes all wage earn-
ers and salaried employees but excludes the self-employed, civil servants, and the military.
It furthermore includes all unemployed who receive unemployment compensation.18 The
data is collected directly on the employer level by the Federal Institute of Employment
and provides detailed employment histories of 460,000 individuals in West Germany and,
after 1992, 110,000 in East Germany. For a detailed description of the data set see Bender
et al. (2000). The basis of my analysis are all individuals aged 15 to 64. I construct the
relative skill shares in the local labour force in each of West Germany’s 204 labour market
regions both by education level and occupation for each year between 1996 and 2001.
In the IAB data I am not able to distinguish ethnic German immigrants from native Ger-
mans so that part of the observed change in the employment/labour force rate and the
log wages in a locality could be simply due to composition effects through newly enter-
ing immigrants. Since the ethnic German immigrants’ labour market outcomes one year
after arrival are substantially worse than they are for the resident population (Bauer and
Zimmermann, 1997), their inclusion in the calculation of average labour market outcomes
17The 1999 Microcensus is the first Microcensus that asks German citizens for their year of arrival inGermany which is why I cannot use earlier Microcensuses for the years 1996 and 1997. Furthermore, thereference week in the German Microcensuses is usually the last week of April so that I cannot use theMicrocensus in say 2001 to calculate the skill shares in 2001.
18In 2001, 77.2% of all workers in the German economy were covered by social security and 78% ofunemployed individuals in West Germany received official unemployment compensation - mostly eitherunemployment benefits (Arbeitslosengeld) or unemployment assistance (Arbeitslosenhilfe) - and are hencerecorded in the IAB data (Bundesagentur fur Arbeit, 2004). The data set does not provide information onthe out of labour force population and those individuals which are currently actively looking for a job buthave not yet paid into the social security system.
17
would lead to a downward bias of the true change in labour market outcomes for the resi-
dent population. For that reason, I make use of the longitudinal dimension of my data set
and restrict the sample to those individuals that were already observed in the data before
1996 when constructing the skill-group specific average employment/labour force rates
and wages.19
These employment/labour force rates and wages are obtained by regressing separately for
each year and skill group the individual level outcomes, either an employment indicator
or log wages, on a set of observables, including a cubic of potential experience, a vector
of region fixed effects, and a set of education (for the occupation-based regressions) and
occupation (for the education-based regressions) group fixed effects. In addition, I include
sixteen country/region of origin dummies as well as a gender dummy when I am pooling
native Germans and resident foreign nationals as well as men and women to construct
labour market outcomes for the overall population.20 In each case, I use the estimated
coefficients on the region dummies as the dependent variables in the regressions of Equa-
tions 1 and 2. They reflect the employment/labour force rate and average log wage in each
locality, adjusted for observable differences in experience, gender, origin, and educational
(occupational) composition within each occupation (education) group across local labour
markets. All outcomes are constructed for the 31st of December of each year.21
For my analysis, the IAB sample has two major advantages compared to other data
sources. First, since I am dealing with administrative data which is used to calculate
19Although this procedure effectively excludes all newly immigrating ethnic Germans from the calcula-tion of average labour market outcomes, it also excludes all those individuals who are starting their first jobbetween 1996 and 2001 or who were self-employed before 1996 and are now entering an employment thatis subject to social security contributions.
20The countries and regions I distinguish are Turkey, former Yugoslavia, Italy, Greece, Poland, the formerSoviet Union, Portugal, Romania, Western Europe, Central & Eastern Europe, Africa, Central & SouthAmerica, North America, Asia, Australia & Oceania and Others.
21I chose the 31st of December to conform with the available data on annual inflows of ethnic Germanimmigrants as well as the reference date used in the official population data of the German Statistical Officewhich I merged with the IAB data.
18
health, pension and unemployment insurance contributions, the precision of the data is
high. In particular the wage data are unlikely to suffer from any measurement error or
reporting bias typical in many survey data sets.22 Second, the sample size is large and
includes detailed regional identifiers. This is necessary because I look at different sub-
groups of individuals in Germany’s local labour markets. Even with an annual sample
size of 460,000 observations, cell sizes quickly become rather small when disaggregating
the labour force by locality, gender, education levels and occupations.
4.4 Federal Statistical Office
Finally, I use county level population data provided by Germany’s Federal Statistical Of-
fice to calculate overall ethnic German immigrant inflow rates into each county, which
are needed in order to evaluate the effectiveness of the Assigned Place of Residence Act.
From the population data, I also construct local growth rates of both the German and the
foreign population, which I use to investigate whether there is evidence of out-migration
in response to the inflow of ethnic German immigrants (see Section 6.2).
5 Descriptive Evidence
5.1 Definition of Skill Groups and Labour Market Regions
The theoretical model suggests that immigration affects relative labour market outcomes
by changing the relative skill shares in the local economy. I differentiate skill groups in
two ways. First, I use the reported educational attainment of an individual, distinguishing
three different groups: low, intermediate and high. People with low education are individ-
uals without an apprenticeship, people with intermediate education are individuals with
an apprenticeship and people with high education are individuals with college education.
22Wage records in the IAB data sample are top coded at the social security contribution ceiling. I imputethose wages by first estimating a tobit model and then adding a random error term to the predicted value ofeach censored observation ensuring that the imputed wage lies above the threshold (see Gartner, 2004 fordetails).
19
Apprenticeships are a crucial component of Germany’s educational system and more than
two thirds of all Germans have completed one in 2001. Individuals usually enter appren-
ticeships immediately after leaving school. They typically consist of two to four years on
the job training with complementary class room teaching one day per week. In terms of
future income, apprenticeships are a more important determinant than the actual number
of years an individual went to school. For instance, the average daily wage of German in-
dividuals without an apprenticeship in West Germany in 2001 is e46.5 if they do not have
A-levels, and only marginally higher at e47.1 if they do. For that reason, I choose them
as the prime indicator of an individual’s skill level in terms of educational attainment.
Second, as an alternative and to check the robustness of the empirical results, I define skill
groups along five different occupation lines (see also Card, 2001): I. farmers, labourers
and transport workers, II. operatives, craft workers, III. service workers, IV. managers,
sales workers, and V. professional & technical workers. For the immigrant population
these occupations refer to the last occupation in the country of origin. The motivation
for this disaggregation by occupation is that the reported level of education an immigrant
obtained in his or her country of origin does not necessarily correspond well to the cor-
responding level of education in the host country.23 Natives and immigrants in the same
occupation group might therefore better reflect comparable skill levels.24
Table 1 provides some descriptive statistics on the overall ethnic German population im-
migrating in each year between 1996 to 2001. In 1996, 177,751 ethnic German immi-
23However, because of their cultural links with Germany, ethnic German immigrants are presumably ina better position to appropriately respond to questions in the Microcensus on their educational attainmentthan, for instance, foreign nationals.
24Borjas (2003) defines skill groups in terms of education and work experience, arguing that individ-uals with similar education but different experience in the labour market are imperfect substitutes in theproduction process. Due to relatively small sample sizes in the German Microcensus from which I take theinformation on educational attainment and the unavailability of cross-tabulations of occupational attainmentby age group, it is unfortunately not possible to extend my analysis in this direction and allow for imperfectsubstitutability across age groups. Similarly, since I cannot distinguish ethnic German immigrants fromnative Germans in my data, I am not able to allow for imperfect substitutability between natives and im-migrants within the same skill group as suggested in two recent studies by Ottaviano and Peri (2006) andManacorda et al. (2006) for the U.S. and the UK, respectively.
20
Table 1: Descriptive statistics of ethnic German immigrants, 1996 to 2001
Year Overall1996 1997 1998 1999 2000 2001 1996 - 2001
Overall inflow 177,751 134,419 103,080 104,916 95,615 98,484 714,265
Men 85,918 65,010 49,664 50,456 46,145 47,379 344,572Women 91,833 69,409 53,416 54,460 49,470 51,105 369,693
Mean % inflow rate* 0.19 0.17 0.13 0.12 0.12 0.12 0.84(standard deviation) (0.10) (0.08) (0.05) (0.05) (0.05) (0.04) (0.33)
% Labour force 53.6 53.7 55.0 55.6 56.6 57.3 55.0
% Age < 15 27.6 26.2 25.5 24.2 23.3 22.6 25.3% Age 15-64 65.9 66.5 67.8 69.0 70.1 71.1 68.0% Age > 64 6.5 7.3 6.7 6.8 6.6 6.3 6.7
% Occupation I 28.3 28.9 27.3 27.5 28.4 26.1 27.9% Occupation II 29.0 28.6 31.0 30.3 30.5 31.5 30.0% Occupation III 18.7 18.3 17.9 17.7 18.4 18.7 18.3% Occupation IV 4.4 4.8 4.7 5.5 5.3 4.8 4.9% Occupation V 19.6 19.4 19.0 18.9 17.4 18.8 18.9
% Low education 47.2 48.8 36.3 43.6 34.4 45.3 43.3% Intermediate education 43.8 42.9 49.3 46.1 53.1 46.4 46.4% High education 9.0 8.3 14.4 10.2 12.5 8.4 10.2
Source: Bundesverwaltungsamt* Mean inflow rate based on 148 West German labour market regions.Note: The educational attainment composition is obtained from the German Microcensuses 1999, 2001 and 2002.Labour force participation and occupation refer to last activity in country of origin and is reported upon arrival.
grants came to Germany. This number gradually declined to 95,615 in 2000 and then in-
creased again slightly to 98,484 in 2001. Overall, over the period 1996 to 2001, 714,265
ethnic German immigrants came to Germany, which corresponds to an average inflow
rate relative to the resident population of 0.84% using the 148 West German labour mar-
ket regions for which I was able to obtain the relevant data. From the descriptives on the
age and occupational composition of the ethnic German immigrants we can see that the
immigrant cohorts remain relatively homogenous over time. There is a slight increase in
the labour force participation in the home country before immigration, which rises from
53.6% in 1996 to 57.3% in 2001. Furthermore, the immigrant cohorts became slightly
older over time, with 22.6% being less than 15 years old, 71.1% of working-age 15 to 64,
and 6.3% older than 64 in 2001. The structure of the occupational composition, which
is reported upon arrival in Germany, did not change substantially over time. There is a
slight decrease in the number of immigrants working in low skill occupation group I from
21
Table 2: Summary statistics for West German labour market regions. Meansand standard deviations
Year Change1996 1997 1998 1999 2000 2001 1996 - 2001
Overall population 315,791 316,413 316,776 317,788 318,762 320,210 1.9%(382,216) (382,306) (382,297) (383,852) (386,969) (388,474) (2.6%)
Working-age pop. (15-64) 214,304 214,368 214,383 214,263 214,049 214,358 0.5%(266,845) (266,338) (265,945) (265,956) (266,289) (266,984) (2.9%)
Foreign immi. share (in %) 10.5 10.4 10.2 10.2 10.1 10.1 −0.3(4.2) (4.0) (4.0) (4.0) (4.0) (3.9) (1.1)
Labour market outcomes:
Lf/pop rate 53.0 51.9 52.8 53.0 53.4 53.3 0.1(7.3) (7.3) (7.5) (7.7) (7.9) (8.2) (1.9)
Empl/pop rate 47.4 46.9 47.6 48.6 49.4 49.6 1.2(6.9) (7.0) (7.2) (7.5) (7.8) (8.1) (2.1)
Unempl/pop rate 5.6 5.0 5.2 4.3 4.1 3.8 −1.1(1.2) (1.2) (1.3) (1.1) (1.1) (1.1) (0.7)
Empl/lf rate 89.4 90.3 90.2 91.8 92.4 93.0 2.2(2.2) (2.3) (2.4) (2.1) (2.2) (2.1) (1.2)
Mean daily wage (in e) 75.1 74.7 75.3 76.1 76.1 76.7 1.7%(6.5) (6.7) (6.8) (7.0) (7.0) (7.2) (2.0%)
Socioeconomic characteristics:
% Low education 25.1 24.9 24.7 24.6 24.4 24.0 −1.3(3.2) (3.0) (2.9) (2.9) (2.9) (2.8) (1.5)
% Intermediate education 67.8 68.0 67.5 67.2 67.2 67.2 −0.4(3.1) (3.1) (3.1) (3.4) (3.5) (3.6) (2.0)
% High education 7.1 7.0 7.9 8.2 8.4 8.8 1.7(3.1) (3.2) (3.4) (3.6) (3.7) (3.9) (1.0)
% Occupation I 18.8 18.5 18.2 17.9 17.6 17.0 −2.1(3.4) (3.4) (3.4) (3.4) (3.5) (3.5) (1.0)
% Occupation II 23.0 22.9 22.7 22.2 21.9 21.4 −1.8(4.9) (5.0) (5.1) (5.2) (5.3) (5.4) (1.2)
% Occupation III 33.4 34.0 33.9 34.5 35.1 35.9 3.0(3.7) (3.8) (3.8) (3.8) (3.9) (4.0) (1.4)
% Occupation IV 14.7 14.9 14.8 14.9 15.0 15.2 0.5(2.7) (2.8) (2.8) (3.0) (3.0) (3.1) (1.0)
% Occupation V 10.1 9.7 10.3 10.4 10.4 10.5 0.4(2.6) (2.6) (2.6) (2.7) (2.7) (2.8) (0.8)
% Female 51.2 51.2 51.2 51.2 51.2 51.1 −0.1(0.5) (0.5) (0.5) (0.5) (0.5) (0.5) (0.1)
Mean age 38.2 38.3 38.4 38.5 38.6 38.7 0.7(0.9) (0.9) (0.8) (0.8) (0.8) (0.7) (0.5)
Source: IAB sample, Statistical OfficeNotes: For the labour market outcomes and the socioeconomic characteristics I only consider the working-age popu-lation aged 15-64. Employment and unemployment refers to individuals subject to social security contributions. Basisof this table are West Germany’s 204 labour market regions.
28.3% in 1996 to 26.1% in 2001 and a corresponding increase in occupation group II from
29.0% to 31.5%. There is, however, some variation in the educational attainment of the
arriving immigrant cohorts. For instance the share of ethnic German immigrants with low
education ranges from 34.4% in 2000 to 48.8% in 1997 and the share of those with high
education from 8.3% in 1997 to 14.4% in 1998.
The primary regional unit in my analysis is the West German labour market region. These
22
regions are aggregates of counties which are the original regional units at which I ob-
serve ethnic German inflows. The aggregations take account of commuter flows so that
labour market regions better reflect separate local labour markets. They comprise on aver-
age around 320,000 individuals (compared to around 225,000 for counties), although this
number varies substantially ranging from 64,000 to 2.7 million. Table 2 provides some
descriptive statistics of the labour market outcomes and socioeconomic characteristics of
the population in West Germany’s 204 labour market regions.
5.2 Labour Market Competition of Resident Workers and Immigrants
The theoretical model predicts that ethnic German immigrants only affect relative labour
market outcomes if their inflow leads to changes in the relative supply of different labour
inputs. This would require the ethnic German immigrant population to differ from the
resident population with respect to their skill distribution.
Comparing the educational attainment of the ethnic German immigrants reported in Ta-
ble 1 with the attainment of the resident population reported in Table 2 shows that more
than 43% of the immigrants have a low education level, compared with only 25% of the
resident population. On the other hand, 46% of the ethnic German immigrants have ob-
tained an intermediate education, compared with about 67% of the resident population.
The shares with high education are similar for both groups at around 10% and 8% respec-
tively.
With regard to the occupational distribution, the differences are similarly pronounced.
Close to 60% of the immigrants worked in low skill occupation groups I and II before
coming to Germany, compared with only about 40% of the resident population. While
they are less likely to have worked in the service (∼ 18% vs. ∼ 34%) and, in particular,
the commercial sector (∼ 5% vs. ∼ 15%), a relatively large fraction previously worked
in high-skill occupation group V (∼ 19% vs. ∼ 10%), for instance as mathematicians,
23
engineers, and teachers.
A more systematic way of measuring the degree of dissimilarity in the occupational dis-
tributions is to compute the following index of congruence for any two groups k and l (see
Welch, 1999):
Ckl = ∑c(qkc− qc)(qlc− qc)/qc√(∑c(qkc− qc)2/qc)(∑c(qlc− qc)2/qc)
where qhc gives the fraction of group h(h = k, l) in occupation c, and qc gives the frac-
tion of the entire labour force in that occupation. The index Ckl equals one if the two
groups have identical occupational distributions, and minus one if they are clustered in
completely different occupations. An index close to one therefore implies a high degree
of competition between the two groups under consideration, a value close to minus one
little competition in the labour market. Table 3 displays the occupational distribution for
different subgroups of the native German population as well as the foreign nationals that
live in Germany in 2001.25 In the bottom row, I report the occupational composition of
the cohorts of ethnic German immigrants that arrived between 1996 and 2001 as reported
upon arrival and shown in the last column of Table 1. The rightmost column presents
the corresponding values of the index of congruence Ckl between recent ethnic German
immigrants and the various subgroups of the native German and foreign population.
The results show that ethnic German immigrants are most similar in their occupational
distribution to native Germans with low education with a calculated index of 0.32. This
index drops to -0.95 for Germans with intermediate education but increases again for
25Note that the corresponding fractions are computed using both employed and unemployed individuals,in the latter case using the last occupation they worked in which are imputed in the IAB data set. Theimplicit assumption is thus that individuals do not switch between occupations which is reasonable in thecase of broadly defined occupation groups. Using both employed and unemployed individuals gives a betterindication of the actual labour supply in each occupation group.
24
Table 3: Occupational distributions and index of congruence
2001 Fraction in occupation group Index ofI II III IV V congruence
Native GermansLow education 24.7 27.8 33.3 10.5 3.6 0.32Intermediate education 15.3 20.7 38.8 17.2 8.0 -0.95High education 1.5 1.0 26.9 18.8 51.9 0.21All 16.1 20.4 36.7 15.9 11.0 -0.63
Foreign NationalsLow education 33.1 37.0 24.0 4.6 13.3 0.57Intermediate education 25.2 29.6 28.8 11.3 5.1 0.51High education 2.9 2.3 26.5 16.1 52.3 0.26All 27.4 31.9 27.3 7.8 5.5 0.63
Ethnic German immigrants 27.9 30.0 18.3 4.9 18.9 1.00
Source: IAB sample, BundesverwaltungsamtNotes: The occupation groups are I: farmers, labourers, transport workers; II: operatives, craft workers;III: service workers; IV: managers, sales workers; V: professional & technical workers. The occupa-tional composition refers to last activity in country of origin of all ethnic German immigrants thatarrived between 1996 and 2001.
highly educated Germans to 0.21. The index of congruence with respect to the over-
all native German population is -0.63, indicating the substantially different occupational
composition compared to the immigrating ethnic Germans. The corresponding index for
the resident foreign nationals in Germany is 0.63, which in turn means that these are quite
similar in their occupational composition to the newly arriving ethnic German immigrants.
Within the group of foreign nationals those with low and intermediate education levels are
most similar with indices of 0.57 and 0.51 respectively. Based on these calculations, the
immigrant inflows between 1996 and 2001 are likely to have exerted supply pressure on
the labour markets of particularly the foreign nationals in Germany as well as the less ed-
ucated native Germans. There is also some indication of increased supply pressure for the
highly skilled native labour force. Due to initial occupational downgrading of the more
highly skilled ethnic German immigrants, however, some of this pressure may have been
shifted away towards the lesser skilled resident labour force (Bauer and Zimmermann,
1999).
To conclude, both the educational and occupational composition of the newly arriving
25
ethnic German immigrants differs substantially from the existing skill composition of,
in particular, the native German population and will therefore have affected the relative
factor supplies in the economy.
5.3 Variation in Existing Skill Compositions
As described in Section 3.2, the primary source of variation in my empirical analysis arises
from differences in the existing skill composition of the labour force across local labour
markets. As the summary statistics for West Germany’s 204 labour market regions in Ta-
ble 2 indicate, there is considerable variation in skill shares both in terms of occupations
and educational attainment. To illustrate this point, I calculate the index of congruence as
defined in the previous section between the existing skill composition in each locality at
the end of 1995 and the skill attainment of the ethnic German immigrants. The map on the
left of Figure 3 shows this index of congruence with respect to occupations for all West
German labour market regions while the map on the right shows the corresponding index
with respect to educational attainment. As before, the index ranges between minus one
and plus one, the former signifying that the local labour force and the immigrants have
entirely different skill compositions and the latter indicating identical skill compositions.
Both maps underline the substantial variation in existing local skill compositions across
West Germany and the consequential variation in differences relative to the skills of the
arriving ethnic German immigrants. These differences across regions give rise to different
labour market effects even if all regions are exposed to homogenous immigrant inflows in
terms of relative size and skill composition.
To give an example, the lowest share of individuals with low education in a locality is
18.3% (county Nordfriesland in Schleswig-Holstein) while the highest share is 41.5%
(county Zollernalbkreis in Baden-Wurttemberg). Using Equation 3 and given an average
overall ethnic German inflow rate between 1996 and 2001 of i = 0.84% of which v=43.3%
had only low education (compare Table 1), the corresponding percentage change in the
26
Figure 3: Index of congruence across West German labour markets
Cologne
Munich
Hamburg
Berlin
Frankf urt
Occupation Index
-1 - -0.50
-0.49 - 0.00
0.01 - 0.50
0.51 - 1.00
Cologne
Munich
Hamburg
Berlin
Frankf urt
Education Index
-1.00 - -0.50
-0.49 - 0.00
0.01 - 0.50
0.51 - 1.00
share of individuals with low education is then 0.04% for the region with the highest ini-
tial share, and 1.1% for the region with the lowest initial share. Similarly, for high skill
individuals, the lowest share in my labour market regions is 1.9% (county Cochem-Zell
in Rhineland-Palatinate) while the highest is 12.9% (area of Darmstadt and Darmstadt-
Dieburg in Hesse). With 10.2% of the ethnic German immigrants being college educated,
this leads to a percentage change in the corresponding skill share of -0.17% for the ini-
tially high-skill, and 3.6% for the initially low-skill local labour market.
The variation in existing skill shares with respect to occupation groups is similarly pro-
nounced. For instance at the end of 1995, the share of individuals belonging to occupation
group I ranges from 13.6% (county Calw in Baden-Wurttemberg) to 29.1% (county Holz-
minden in Lower Saxony) while the share belonging to high-skill occupation group V
ranges from 3.0% (county Cochem-Zell in Rhineland-Palatinate) to 17.9% (county Lev-
27
erkusen in North Rhine-Westphalia). It is this variation in the existing skill compositions
across German labour markets that identifies the effect of ethnic German inflows on local
labour market outcomes.
5.4 Exogeneity of Allocation
The validity of my instrumental variable based on the ethnic German immigrant inflows
relies upon the effectiveness of the Assigned Place of Residence Act and the exogeneity
of the immigrants’ allocation by the authorities with regard to transitory local demand
conditions. Since the main allocation criterion was the proximity of family members and
labour market skills did not feature in any significant way in the allocation process, the ex-
ogeneity requirement is likely to be satisfied. In fact, if family ties were the only criterion
by which immigrants would choose their place of residence themselves, one would not
require the government allocation policy in order to maintain the exogeneity assumption
with regard to local labour demand shocks. However, local labour market conditions are
likely to have played a role in the choice of place of residence before the introduction of
the new legislation in 1996, as suggested by Figure 4.
Figure 4 shows the variation of ethnic German immigrant inflow rates in all West German
counties before the introduction of the new legislation in 1996 and for the counties where
the law was implemented thereafter. There is a significant reduction in the variation of the
regional inflow rates after the introduction of the new legislation. This reduction indicates
that the new allocation policy has indeed been effective in altering the direction of ethnic
German immigrant inflows and ensuring a more even distribution across Germany. It also
points towards the existence of a few particularly attractive destinations before 1996.
There are several potential reasons for the remaining variation after 1996 shown in Figure
4. First, the quotas for each federal state and a large number of counties have not been
adjusted to changes in their corresponding populations after they were originally set. In
28
Figure 4: Variation in the ethnic German immigrant inflow rate, 1989 to 2001
−.0
050
.005
.01
.015
Dev
iatio
ns fr
om m
ean
1990 1992 1994 1996 1998 2000
Notes: Values depicted are deviations from the mean ethnic German inflow rate in each year. The inflowrates are calculated as the number of allocated ethnic German immigrants divided by the overall populationin the county at the end of the previous year. The sample size is, 85 for 1989, 145 for 1990/1991, 204 in1992-1994, 230 in 1995, 122 in 1996 and 168 in 1997-2001. From 1996 onwards only counties in statesthat implemented the Assigned Place of Residence Act are depicted.
addition, when the state quotas were set in 1993, they were not exclusively based on the
resident population but also on the strength of the economy of each state so that some
states (and thus the counties they comprise) might receive higher relative inflows than
others. I control for these differences in my empirical estimations by the inclusion of
region fixed effects. Another reason for the observed differences in relative inflows are
different allocation procedures. For instance, in North Rhine-Westphalia the geographical
area of each county features as an additional factor in determining the number of immi-
grants allocated and in Lower Saxony some counties which received a disproportionate
number of ethnic Germans in the early 1990s were exempted from additional allocations
for some years after 1996.
Trying to achieve an even distribution while giving as much consideration as possible to
the proximity of family members are two not always reconcilable objectives. In some
29
cases ethnic German immigrants could not be allocated to their desired destinations even
if they had relatives living there because those regions had already met their quotas. In
these cases they were typically allocated to an alternative region close by. This prece-
dence of an even distribution over family ties could potentially be quite important for two
reasons. First, if every arriving ethnic German immigrant was guaranteed to be allocated
to the region where his or her relatives lived, then there would in theory be scope for a
selective choice of the time of arrival in Germany in order to take advantage of particu-
larly good local demand shocks. However, in practice, independent of the precedence of
an even distribution, current labour market conditions did not seem to have played any
significant role in determining an immigrant’s time of arrival because the economic ben-
efits of moving were typically not contingent upon getting a paid job in Germany upon
arrival.26 On the aggregate level, there is no evidence that ethnic German immigration is
positively related to overall labour market conditions. On the contrary, as Tables 1 and
2 show, while both employment and wage rates in Germany increased steadily between
1996 and 2001, ethnic German inflows gradually decreased. To investigate this issue in
more detail, I regress the annual inflow rates into each region on the employment/labour
force rate and the wage level at the beginning of each year, including both year and region
fixed effects. In the absence of county quotas, and if immigrants were certain about which
area they would be allocated to and were timing their arrival based on the labour market
situation in that area at the beginning of each year, one would expect to find a positive
correlation between initial labour market conditions and immigrant inflows. Both coeffi-
cient estimates of these regressions are virtually zero and statistically not significant with t
statistics of -0.03 and 0.58 respectively.27 Whether the absence of any correlation is due to
government authorities strictly adhering to the set quotas and not allowing relatively more
immigrants to move into regions with particularly good current labour market conditions,
or immigrants not timing their arrival accordingly cannot be directly deduced from these
26According the government authorities it seemed to be predominantly factors in the country of originthat determined the actual timing of immigration to Germany.
27The point estimate on the employment/labour force rate is −0.71 ·10−4 with a robust standard error of24.4 ·10−4 while the estimate on the average wage level is 0.19 ·10−4 with a standard error of 0.33 ·10−4.
30
results. To answer that question I would require information on the number of immigrants
that arrived in Germany each year but were not allocated to their preferred destination.
If these numbers were positively related to current labour market conditions, this would
point towards a selective timing of immigration. What the results show, however, is that
local labour market conditions at the beginning of a year did not affect the size of relative
inflows into each area.28 The second potential problem that could arise if there was no
precedence of quotas over family ties is that, theoretically, relatives could move to those
areas that are particularly attractive before the immigration of the ethnic German occurs
and through this channel allow an endogenous self-selection of the immigrant. However,
even in that case, as long as the selective migration of relatives is based on permanent
rather than transitory features of the selected labour market region, I am able to control
for such behaviour by including region fixed effects in the empirical estimations.
One way to investigate whether the allocation decision has indeed been exogenous with
respect to individual skill characteristics as suggested by the overwhelming importance of
family ties for the allocation decision is to compare the age distribution of the ethnic Ger-
man immigrants that were allocated to each federal state. These distributions are recorded
at the central admission centre and reported in Table 4. If immigrants were exogenously
allocated with respect to their individual characteristics, one would not expect there to
be significant differences in their age distribution across states. As shown in Table 4, the
age distributions across states are indeed very similar. As a reference point, I show the
standard deviation of each age group’s share of the overall resident population across the
same states at the end of 1995 in the last column. Apart from the 15 to 24 year-olds, the
standard deviation of the age group shares of the allocated ethnic German immigrants is
28If relative labour market conditions for different skill groups lead to selective relative timing of arrivalby these skill groups, then this could potentially be problematic. For example, if there are good conditionsfor low-skill workers in a locality relative to those for high-skill workers, this could lead to an advancementof immigration by low-skill workers and a postponement by high-skill workers, thus changing the compo-sition (rather than the size) of the arriving immigrant labour force. For the construction of my instrumentalvariable I assume that the skill composition of the arriving ethnic German immigrants in each locality isidentical.
31
Table 4: Age distribution of allocated ethnic German immigrants, 1996 to 2001
Agegroup SH HA LS BR NW HE RP BW BA SA STDEV STDEV all
0 - 14 25.9 24.2 26.4 26.1 25.9 25.8 25.6 25.0 25.0 24.8 0.7 1.2
15 - 24 18.7 19.7 19.2 18.9 19.3 18.6 19.1 18.9 19.0 18.9 0.3 0.3
25 - 34 15.3 15.0 14.9 15.3 14.9 15.3 15.0 14.8 14.9 15.3 0.2 0.7
35 - 44 18.2 17.8 18.0 17.5 17.7 17.8 17.4 17.8 17.7 17.9 0.2 0.5
45 - 55 9.1 10.1 8.8 9.2 9.0 8.9 9.7 9.5 9.5 9.8 0.4 0.6
55 - 64 6.4 7.1 6.6 6.8 6.6 6.7 6.6 7.0 7.2 7.0 0.3 0.4
> 64 6.4 6.2 6.2 6.3 6.6 6.8 6.6 7.1 6.7 6.3 0.3 0.8
Notes: West Germany’s 10 federal states are: Schleswig-Holstein (SH), Hamburg (HA), Lower Saxony (LS), Bremen (BR),North Rhine-Westphalia (NW), Hesse (HE), Rhineland-Palatinate (RP), Baden-Wurttemberg (BW), Bavaria (BA) and Saar-land (SA).
substantially lower than the corresponding standard deviation in the overall population in
all age groups. In particular the shares of the groups aged 25 to 34 and 35 to 44, which
represent a large part of the working population and are therefore most relevant for this
analysis, are very similar across states. A regression of the age group shares of the immi-
grant population allocated to each state between 1996 and 2001 on the existing share at the
end of 1995 and a set of age group fixed effects gives an estimate of −0.03 with a robust
standard error of 0.12.29 Hence there is no evidence that for instance young ethnic Ger-
man immigrants have been allocated to states that are generally more attractive to young
people. Overall the figures suggest that there has been an exogenous allocation of ethnic
German immigrants to each federal state with respect to their individual characteristics.
Since the allocation to each state follows similar administrative processes and decision
criteria as the subsequent allocation to different counties, the results in Table 4 can be
regarded as indicative of an exogenous allocation within states to different counties.
29Similarly, regressing annual age group shares on existing age group shares of the resident population aswell as interactions of age group and year fixed effects gives a statistically not significant estimate of−0.01with a robust standard error of 0.07.
32
6 Empirical Results
6.1 Employment and Wage Effects
Turning to the estimation results, Table 5 presents estimates of the effect of changes in
skill-specific local labour force shares on the employment/labour force rate of the resi-
dent population. I estimate the empirical model in Equation 1 first by OLS and then using
the predicted skill-specific ethnic German inflow rate as described in Section 3.1 to instru-
ment the potentially endogenous change of the skill shares in a locality. I report results
for skill groups based on occupations in the upper panel and for skill groups based on
educational attainment in the lower panel. The dependent variable in each regression is
the regression-adjusted employment/labour force rate of the local labour force, thus con-
trolling for differences in individual characteristics across labour markets. The estimates
in columns (1) and (2) are based on all 148 West German labour market regions for which
data on ethnic German inflows are available while in columns (3) and (4) the sample is re-
stricted to those 112 regions that formally implemented the Assigned Place of Residence
Act. The reason why the inclusion of labour market regions in states that have not for-
mally implemented the legislation could be of interest is that even in those states the main
criterion for the actual allocations were family ties, in which case the immigrant inflows
would also be exogenous to unobserved labour demand shocks and provide additional
observations for the estimations. However, endogenous allocations by the authorities as
well as self-selection by immigrants within these states continues to be a possibility, so
that the results from this specification are likely to remain upward biased.
Looking at the OLS results for all individuals reported in the first row in columns (1) and
(3) of the upper panel first, we see a significant negative effect of an increase in the relative
occupation share in a locality on the overall employment/labour force rate. The estimated
coefficients of -0.125 and -0.126 imply that a 10% increase in the relative occupation
share induced by additionally employed individuals reduces the employment/labour force
33
Table 5: Impact of changes in relative factor shares on the employment/labour force rate
All regions Restricted regionsOLS IV OLS IV(1) (2) (3) (4)
Occupation groups
All -0.125*** -0.026 -0.126*** -0.353**(.011) (.306) (.013) (.168)
[1.44] [3.13]
All unweighted -0.120*** 0.127 -0.121*** -0.374**(.012) (.451) (.012) (.189)
[1.38] [2.98]
All aged 25-54 -0.118*** 0.109 -0.122*** -0.211(.011) (.264) (.012) (.150)
[1.80] [3.17]
Germans only -0.125*** -0.090 -0.122*** -0.327**(.011) (.222) (.012) (.155)
[1.84] [3.40]
Observations 4440 4440 3185 3185
Education groups
All -0.069*** -0.381* -0.074*** -0.482*(.017) (.198) (.019) (.288)
[2.94] [2.66]
All unweighted -0.070*** -0.348 -0.065*** -0.248*(.015) (.212) (.018) (.132)
[2.95] [3.21]
All aged 25-54 -0.065*** -0.234 -0.067*** -0.416(.020) (.235) (.020) (.258)
[2.49] [2.74]
Germans only -0.079*** -0.313* -0.083*** -0.425(.018) (.181) (.019) (.267)
[3.24] [2.58]
Observations 2664 2664 1911 1911
Notes: Entries are the estimated coefficients on the change in the log factor shares ∆log f jrt . Thedependent variable is the annual change in the skill-specific employment/labour force rate. Allestimations include five occupation and three education groups respectively. Columns 1 and 2use all 148 West German labour market regions for which data is available, columns 3 and 4 onlythose 112 that actually implemented the law (see Table B-1 in Appendix B). Employment/labourforce rates are based on individuals already in the data at the end of 1995. Additional covariatesare a full set of interactions of skill and year fixed effects as well as region and year fixed effects.Employment/labour force rates are adjusted for differences in individual specific characteristicsacross labour markets. Robust standard errors are reported in parentheses and are clustered at theskill-specific regional level. For the IV estimates, the t-stat of the instrument from the first stageregression is reported in square brackets. Regressions are weighted by the inverse of the standarderrors of the region fixed effects taken from the regressions to obtain adjusted outcomes. A (*)denotes statistical significance at the 10% level, a (**) at the 5% level and a (***) at the 1% level.
rate of the resident population by 1.25 and 1.26 percentage points respectively.30
30Note that in order to facilitate the calculation of regression-adjusted employment/labour force rates Iuse the employment/labour force rate in levels in my estimations rather than in logs as suggested by thetheoretical model in Section 3.1. One can translate the coefficients in my tables for the effects on theemployment/labour force rate into estimates of β1 by dividing them by the average employment/labour
34
In the presence of unobserved transitory local demand shocks, the OLS estimates of Equa-
tion 1 will be upward biased since such shocks attract workers into a particular skill group
while at the same time improving employment opportunities. I therefore instrument the
changes in the relative skill shares with the occupation-specific ethnic German inflow rate.
The corresponding estimates are reported in column (2) and (4). While the coefficient for
the specification based on all labour market regions is small and statistically not signif-
icant due to a weak first stage with a t statistic for the instrument of only 1.44, restrict-
ing the sample to those regions that did formally implement the legislation increases the
strength of the instrument and reduces the estimate to -0.353, which is significant at the
5% level (column 4). Since, as explained in Section 3.1, ethnic German immigrants can
only appear in the data and hence enter the numerator of the relative local skill share by
becoming employed, the estimated coefficients can be directly interpreted as a displace-
ment effect: for every 10 ethnic German immigrants finding employment, 3.5 resident
workers accordingly lose their job (or do not find one when they otherwise would have).
The increase in magnitude of this estimate by a factor of around 3 compared to the OLS
results points towards the existence of unobserved skill-specific local demand shocks that
attract workers into the labour force as well as lead to favourable changes in local labour
market outcomes.
The first row of the lower panel of Table 5 reports results for the same regression but this
time after defining skill groups according to the educational attainment of an individual.
While the OLS results in columns (1) and (3) suggest that an increase in the relative skill
share through additionally employed individuals by 10% reduces the employment/labour
force rate of the resident labour force by 0.69 and 0.74 percentage points respectively,
this effect increases by a factor of 5.5 and 6.5 respectively, to 3.81 and 4.82 percent-
age points once I instrument for the potentially endogenous change in the relative skill
force rates of all individuals (0.91).
35
shares. Although only marginally significant at the 10% level, the point estimates of the
IV regressions in column (2) and (4) suggest a similar magnitude as the one found when
distinguishing between different occupation groups. Moreover, the fact that the IV esti-
mates increase in magnitude when moving from all 148 regions to the restricted sample
of 112 regions indicates that, in the former case, there may be some positive correlation
remaining between the ethnic German inflows and unobserved demand shocks in those
areas where the law has not been implemented so that the estimated coefficient continues
to be upward biased. The implied displacement effects of 3.81 and 4.82 workers for every
10 ethnic Germans finding employment seem relatively large. However, since, based on
information from the German Microcensus, only between 30% and 40% of working age
ethnic German immigrants find a job in the first year after arrival, and absolute inflows
on the local level have been relatively moderate, the actual number of displaced native
German and foreign workers has been quite small.31
The remaining rows of Table 5 show estimates of β1 for a number of alternative specifica-
tions in order to test the robustness of the results. In the second row of each panel, I report
the unweighted regression results for both the OLS and IV estimations. All estimates are
similar in magnitude to their counterparts in the weighted regressions apart from the IV
result based on education groups for the restricted set of regions reported in column (4)
which is somewhat smaller with a point estimate of -0.248. Since the data have some
shortcomings in terms of capturing movements into and out of the labour force, I estimate
my model separately for individuals aged 25 to 54 for which these movements are less of
an option to adjust to changing labour market conditions. The corresponding results are
reported in the third row of each panel. Although statistically not significant, the point
estimates indicate a slightly smaller magnitude than the one found when using all indi-
viduals as reported in the first row of each panel. Finally, I investigate whether there are
31Multiplying the estimated coefficients by the share of immigrants that find employment within the firstyear of arrival will provide an estimate of how a general inflow of immigrants into the labour force, whetheremployed or unemployed, affects labour market outcomes.
36
different effects for the native German population compared to foreign nationals living in
Germany which make up about 10% of the labour force. Due to the limited sample size
for the latter group in my region/skill cells, estimating separately for them is not viable.
However, I can estimate separately for native Germans and compare the results with those
obtained when using all individuals to get at least an indication of whether the effect on
foreign nationals is likely to be larger or smaller than the one on Germans. The last row
of each panel in Table 5 reports the results for the effect on the employment/labour force
rate of the native German population only. Compared to the estimates for the overall
population reported in the first row, the estimated effects tend to be smaller both in the
regressions based on occupations and the ones based on educational attainment. In the
first case, using the restricted set of labour market regions leads to a significant estimate
of -0.327 (column 4) compared to -0.353 when using the entire population, both Germans
and foreign immigrants. Similarly, the estimate based on education groups decreases from
-0.482 for the overall population to -0.425 for the German population, although this esti-
mate is not statistically significant at conventional levels.
Turning towards the impact of changes in relative skill shares on wages, the upper panel in
Table 6 reports the results for the coefficient β2 in Equation 2 when, as before, skill groups
are defined by occupation, whereas the lower panel reports the results when they are de-
fined by education. The OLS estimates of β2 for the wages of all individuals reported in
the first row of Table 6 in column (1) are -0.049 for the occupation and -0.058 for the edu-
cation regressions. These imply that a 10% increase in the relative skill share in a locality
through additionally employed individuals decreases relative wages by 0.49% and 0.58%
respectively. The IV results on the other hand do not show any negative effect of ethnic
German immigrant inflows on the average wage rate both in the specification based on
all 148 labour market regions and the one using only those 112 regions that implemented
the Assigned Place of Residence Act. All estimates are statistically not significant and in
most cases close to zero. The point estimates in the preferred specification in column (4)
37
Table 6: Impact of changes in relative factor shares on log daily wages
All regions Restricted regionsOLS IV OLS IV(1) (2) (3) (4)
Occupation groups
All -0.049*** -0.174 -0.068*** -0.120(.014) (.562) (.015) (.188)
[1.14] [2.69]
All unweighted -0.042*** 0.457 -0.061*** -0.028(.015) (.637) (.015) (.182)
[1.38] [2.98]
All aged 25-54 -0.053*** -0.641 -0.069*** -0.277(.016) (.584) (.016) (.214)
[1.62] [2.71]
Germans only -0.048*** -0.143 -0.066*** -0.197(.014) (.474) (.015) (.192)
[1.33] [2.85]
Observations 4440 4440 3185 3185
Education groups
All -0.058** 0.198 -0.060*** 0.301(.026) (.133) (.022) (.316)
[3.53] [2.08]
All unweighted -0.043** 0.380* -0.071*** 0.084(.021) (.209) (.021) (.130)
[2.95] [3.21]
All aged 25-54 -0.045 -0.019 -0.054** 0.151(.028) (.244) (.022) (.254)
[2.59] [2.33]
Germans only -0.046* 0.298** -0.059*** 0.350(.025) (.115) (.021) (.329)
[4.54] [2.03]
Observations 2664 2664 1911 1911
Notes: Entries are the estimated coefficients on the change in the log factor shares ∆log f jrt .The dependent variable is the annual change in the skill-specific average log daily wage ofall full-time employees. All estimations include five occupation and three education groupsrespectively. Columns 1 and 2 use all 148 West German labour market regions for which datais available, columns 3 and 4 only those 112 that actually implemented the law (see TableB-1 in Appendix B). Average log wages are based on individuals already in the data at theend of 1995. Additional covariates are a full set of interactions of skill and year fixed effectsas well as region and year fixed effects. Average log wages are adjusted for differences inindividual specific characteristics across labour markets. Robust standard errors are reportedin parentheses and are clustered at the skill-specific regional level. For the IV estimates, the t-stat of the instrument from the first stage regression is reported in square brackets. Regressionsare weighted by the inverse of the standard errors of the city fixed effects taken from theregressions to obtain adjusted outcomes. A (*) denotes statistical significance at the 10%level, a (**) at the 5% level and a (***) at the 1% level.
38
are -0.120 with a standard error of 0.188 in the occupation regression and 0.301 with a
standard error of 0.316 in the education regression.
The IV estimates of most of the additional specifications that I estimate and report in
Table 6 are not precisely estimated and inconclusive regarding the effect of ethnic Ger-
man immigrant inflows on relative wages. While the point estimates tend to be negative
in the regressions based on occupation groups, they tend to be positive in the education
based regressions. However, the only cases in which they are statistically significant are
the unweighted specification and the specification for native Germans only based on all
148 regions available (column 2) with estimates of 0.380 and 0.298 respectively. These
positive effects are driven by a large positive impact on wages of German women whereas
the effect on men is very small in magnitude and not significant (see Table D-2 in Ap-
pendix D). As I pointed out before, there remains scope for endogenous self-selection
of immigrants in those regions in which the Assigned Place of Residence Act was not
implemented, which could in principle also lead to a positive coefficient. When I restrict
the sample to the preferred set of 112 regions, the estimates for the education based re-
gressions remain positive but become statistically not significant.
The fact that I do not find any evidence of negative wage effects may be explained by
Germany’s relatively inflexible labour market and, in particular, strong unions and strict
labour market regulations. Although in decline, union coverage is still high at 68% in
2000 (OECD, 2004).32 In addition, wages in Germany are to a large extent set by sector-
level collective wage agreements, leaving little room for wage adjustments on the regional
level. The overall scope for short-term adjustments in the wage structure in Germany in
response to immigrant inflows is therefore limited. This may also explain why I find rela-
tively large adjustments in relative employment levels in my estimations: with rigid wages
and at least some degree of substitutability between the resident workforce and newly ar-
32For comparison, the corresponding figure for the U.S. is 14%.
39
riving immigrants in the production process, an increase in labour supply through immi-
gration leads to an increase in unemployment of the resident population unless it induces
a sufficiently large increase in labour demand. However, as Pischke and Krueger (1998)
point out, constraints and rigidities on the product market are relatively pronounced in
Germany, impacting precisely this demand side of the labour market. For instance, it is
much more difficult to start up a new business in Germany than it is in the U.S. which
contributes to the economy’s sluggishness in creating additional jobs when its population
expands. In fact, total employment in Germany increased by only 1.4% between 1991
and 2001 while the working age population increased by 4.7% (of which around 46% was
due to ethnic German immigrants and 45% due to immigration of foreign nationals).33
This explanation is also supported by the results of a cross-country study carried out by
Angrist and Kugler (2003). Analysing the impact of immigrants on native employment
rates in eighteen European countries, the authors not only find evidence of a substantial
displacement of native workers by immigrants, ranging from 35 to 83 native job losses
for every 100 immigrants in the labour force, but also some clear indication that this ef-
fect is exacerbated by rigidities on the product market, such as high business entry costs,
and reduced flexibility on the labour market, for instance through employment protection,
union coverage, and minimum wages.
As pointed out in Section 3.2, the main source of variation I exploit in the empirical es-
timations are differences in the existing skill compositions across local labour markets.
One concern in this context is that my results may be driven by unobserved trends in skill
region specific labour market outcomes that are correlated with the initial skill share in a
locality. For instance, if for some reason regions with a small initial share of a particular
skill group tend to experience faster declining employment and wage rates than regions
with a large initial share, then even if there was no effect of an immigrant inflow on labour
market outcomes, the empirical estimates would still show a negative effect. This is be-
33Source: Statistical Office and own calculation.
40
cause, as described in Section 3.2, the lower the initial share of a particular skill group
in a locality, the larger will be the percentage change in this share induced by the inflow
of ethnic German immigrants. The observed negative correlation between the percentage
change in the relative skill share and changes in labour market outcomes will in this case,
however, be entirely driven by the underlying correlation between the initial skill share
and future changes in labour market outcomes.
To investigate this issue, I estimate a model relating changes in labour market outcomes
directly to the initial skill shares f jrt−2 in a locality. I use the skill share lagged by two
periods to mimic as closely as possible my previous estimations in which I also used the
skill-specific labour force lagged by two periods to construct the instrumental variable.
The two models for the change in the employment/labour force and wage rate, respec-
tively, are then given by
∆(N jrt/Pjrt) = a jt +art +δ1 f jrt−2 +a jrt
∆ logw jrt = b jt +brt +δ2 f jrt−2 +b jrt ,
where a jt , b jt , art , and brt are, as in the regression models in Equations 1 and 2, interac-
tions of skill group and year fixed effects and region and year fixed effects respectively.
To minimise the influence of any other compounding factors and isolate the effect of ini-
tial skill shares, I estimate these models for the period 1985 to 1987. This is a period of
little immigration to Germany which, at the same time, is sufficiently long after the strong
recession of 1981/82. A significant correlation between the initial skill share f jrt−2 and
changes in labour market outcomes would point towards unobserved skill region specific
trends that are not accounted for in the model set out in Section 3.1.
41
Table 7: Impact of initial skill shares on labour market outcomes, 1985 to 1987
∆(N jrt/Pjrt) ∆ logw jrt
Independent variable Occupation Education Occupation Education
Initial skill share 0.011 -0.003 -0.005 -0.014(.011) (.016) (.016) (.020)
Obs. 1480 888 1480 888R2 0.81 0.76 0.70 0.82
Notes: Entries are the estimated coefficients on the local skill share lagged by two periods, f jrt−2. The dependentvariable is either the annual change in the employment/labour force rate or the annual change in log daily wagesfor the period 1985 to 1987. All estimations include five occupation and three education groups, respectively, andare estimated using West Germany’s 148 labour market regions. Additional covariates are a full set of interactionsof skill and year fixed effects as well as region and year fixed effects. Standard errors are robust and clusteredat the skill-specific regional level. Employment and wage rates are adjusted for differences in individual specificcharacteristics across labour markets (see text). Regressions are weighted by the inverse of the standard errors of theregion fixed effects taken from the regressions to obtain adjusted outcomes. A (*) denotes statistical significance atthe 10% level, a (**) at the 5% level and a (***) at the 1% level.
Table 7 reports the estimates for δ1 and δ2 separately for the regressions based on occu-
pation (columns 1 and 3) and education groups (column 2 and 4). All of the estimated
coefficients on the initial skill share are statistically not significant and close to zero, in-
dicating that the initial skill share is not systematically related to future changes in these
labour market outcomes. For the corresponding results for men and women see Table D-3
in Appendix D. Apart from the effect on women’s wages in the occupation regression,
all estimated gender-specific coefficients are also not significant. Based on these results,
I conclude that unobserved long-term trends correlated with the initial skill shares in a
locality are unlikely to be driving the results of the empirical estimations.
6.2 Migratory Responses
Since the empirical analysis in this paper is based on local labour markets, it is vital to
investigate whether there is evidence for migratory responses of the resident population to
the inflows of ethnic German immigrants. By dissipating the effect of immigration across
the entire economy, one would in that case underestimate the magnitude of the parame-
ters of interest β1 and β2 (see, for instance, Borjas, 2006). Due to Germany’s relatively
inflexible labour market, one would a priori not expect large migration flows in response
42
Table 8: Migratory response of native Germans and foreign nationals to in-flows of ethnic German immigrants
Counties Labour Market RegionsIndependent variable German Foreign German Foreign
Ethnic German inflow rate 1.05 -0.01 1.01 0.08(.19) (.17) (.24) (.25)
Obs. 1380 1380 888 888R2 0.48 0.19 0.42 0.18
Notes: Entries are the estimated coefficients on the ethnic German immigrant inflow rate in models where thedependent variable is either the annual growth rate of the German or the foreign local population in either WestGermany’s 230 counties or 148 labour market regions for which I have information on the annual ethnic Germaninflows between 1996 and 2001. All estimations include a full set of region and year fixed effects.
to increased immigration and previous results seem to confirm this claim (e.g. Pischke
and Velling, 1997). The comparatively generous social security system, with particularly
high and long-lasting unemployment benefits, typically counteracts the incentive to move
to a different location in the face of adverse labour market conditions.34
To formally investigate this issue, I regress the annual growth rate of the German and
foreign population on the annual immigrant inflow rates, including both year and region
fixed effects, the latter to allow for region-specific population growth trends. I estimate at
the county as well as the labour market region level. In the absence of migratory responses
of the resident population to the immigrant inflows, every additional ethnic German im-
migrant moving into a particular county should increase the overall German population
(which includes the ethnic German immigrants) of that county by one while the number
of foreign nationals should remain unchanged. Out-migration of the resident German and
foreign population, on the other hand, would be reflected by coefficient estimates of less
than one and less than zero, respectively. The results from these regressions are shown
in Table 8. As we can see in columns (1) and (3), there is no evidence of native German
out-migration that could dissipate any labour market effects across the economy. Both
34During the 1980s, for instance, the regional disparities of unemployment rates in West Germanywidened substantially while internal migration decreased (see Bauer et al., 2005).
43
estimates are very close to one. Moreover, there is also no evidence that the immigrants
move to areas that are particularly attractive destinations for native Germans, in which
case the coefficient estimate would be greater than one.35 This finding supports the claim
that because of their exogenous allocation to particular counties ethnic German immi-
grants did not self-select into booming local labour markets.
Columns (2) and (4) of Table 8 report the results when I regress the annual growth rate
of foreign nationals in a locality on the ethnic German immigrant inflow rate. As before,
there is no evidence of out-migration of foreign nationals in response to these inflows
which would be reflected by a negative coefficient estimate. Equally important, both for
counties and labour market regions, there is also no indication of a positive relationship
between the flows of ethnic German immigrants and foreign nationals. Both coefficients
are close to zero. Given that foreign nationals are to a large extent free to choose their
place of residence and likely to move to those areas where labour market conditions are
best, one could expect a similar settlement pattern from ethnic German immigrants if they
did indeed choose their places of residence endogenously. In that case the estimates in
Table 8 should show a positive correlation.
Since in the empirical model on which this analysis is based, changes in relative factor
shares are determining the relative wage structure and employment rates, it is instructive
to investigate whether there is evidence of skill-specific out-migration in response to the
inflow of ethnic German immigrants. Following Card and DiNardo (2000), I relate the
annual change in the overall log skill share of a specific skill group in a locality to the
predicted relative immigrant inflow rate for that skill group:
∆log(Pjr/Pr) = a+b(∆I jr/Pjr−1 −∆Ir/Pr−1)+u jr,
35Particularly attractive destinations are in this context regions that experience annual increases in theirGerman population that go beyond their long-term trends.
44
Table 9: Skill-specific migratory response to inflows of ethnic German immi-grants
Counties Labour Market RegionsIndependent variable Occupation Education Occupation Education
Relative inflow rate 1.30 1.65* 1.17 1.74*(.34) (.39) (.42) (.45)
Obs. 6900 4140 4440 2664R2 0.21 0.29 0.30 0.40
Notes: Entries are the estimated coefficients on the relative skill-specific ethnic German immigrant inflow rate. Thedependent variable is the annual change in the log skill share in either West Germany’s 230 counties or 148 labourmarket regions for which I have information on the annual ethnic German inflows between 1996 and 2001. Allestimations include five occupation and three education groups respectively. Additional covariates are a full set ofinteractions of skill and year fixed effects as well as region and year fixed effects. Robust standard errors are reportedin parentheses and are clustered at the skill-specific regional level. Regressions are weighted by the overall skill-specific labour force in each region. A (*) denotes that the parameter is statistically different from 1 at the 10%, a(**) at the 5% and a (***) at the 1% significance level.
where ∆I jr/Pjr−1 is the predicted skill-specific inflow rate of ethnic German immigrants
with skill j in region r and ∆Ir/Pr−1 is the overall inflow rate. If the migratory response
of the resident population fully offsets the skill-specific inflow of immigrants, then the
relative inflow rate will have no effect on the overall skill share and the coefficient b will
be zero. By contrast, in the absence of a differential migratory response of the resident
population in a specific skill group to inflows of ethnic German immigrants into the same
group, the coefficient b will be one. Table 9 shows the results for the parameter b for
both the specification based on occupation groups and the specification based on educa-
tion groups. As before, I estimate at the county as well as the labour market region level.
The results show that there is no indication for any selective out-migration of the resident
population that could offset the changes in relative factor shares induced by the immigrant
arrival. All parameter estimates are larger than 1, with point estimates of 1.30 and 1.17 for
the occupation-based regressions and 1.65 and 1.74 for the education-based regressions.
If at all, there is some evidence that the skill-specific inflow of immigrants leads to an
increase in the relative growth of the corresponding resident population, although only in
the education-based regressions is b statistically different from 1 and that only at the 10%
level.
45
To sum up, overall the results in Table 8 and Table 9 show that there is little evidence of
any out-migration of the resident population, both overall and skill-specific, in response
to ethnic German immigrant inflows. It is therefore unlikely that out-migration has miti-
gated the effect the immigrant inflow has had on the regional wage structure and relative
employment rates.
7 Conclusion
The arrival of ethnic German immigrants and their distribution across local labour mar-
kets by the administration offers a unique natural experiment to investigate the impact of
immigration on labour market outcomes. In this paper, I analyse how these inflows have
affected the employment/labour force rates and relative wages of the resident population
in Germany between 1996 and 2001.
The empirical results show that shifts in the relative supply of different skill groups in a
locality systematically affect the employment/labour force rate of the resident population.
Like previous studies, I find evidence that unobserved skill-specific demand shocks lead
to biased OLS estimates of the effect of these relative supply shifts. Instrumenting them
with the ethnic German inflow rate leads to substantially larger estimates by a factor of
3 to 7. The estimated short-run effects on the overall employment/labour force rate are
relatively stable for both skill definitions, occupations and educational attainment, point-
ing towards a displacement effect of around 0.4 or 4 unemployed resident workers for
every 10 immigrants that find a job. I do not find conclusive evidence of any detrimental
effect on relative wages. When estimating the empirical model for the native German
population alone, excluding resident foreign nationals from the sample, the estimates for
the effect on the employment/labour force rate become smaller in magnitude, suggesting
that resident foreign nationals may be more affected by ethnic German immigrant inflows
46
than the native German population.
While the absence of significant wage effects of immigration is consistent with most of the
existing evidence for Germany, the conclusion that immigrant inflows into a local labour
market have a detrimental effect on the employment/labour force rate stands in contrast to
a number of other studies for Germany, for instance Pischke and Velling (1997) or Bonin
(2005). Both these studies, however, cover a different period, the former the years 1985
to 1989, and the latter the years 1975 to 1997, so that the results are not necessarily com-
parable. In addition, and in contrast to my analysis, the study by Pischke and Velling,
related in that it also uses spatial correlations to identify the immigrant impact, identifies
a medium-run effect of immigration by looking at changes over a four-year period. The
longer time period allows more scope for labour market adjustments through compen-
satory population flows as well as changes in the industry structure and output mix of
the local economy, both channels which would tend to reduce the effect on relative local
labour market outcomes. The fact that German labour markets adjust to immigrant in-
flows through changes in employment rather than wages is potentially due to Germany’s
institutional setting in which strong unions allow relatively little wage flexibility, at least
at the regional level and in the short run. The relatively large magnitude of the displace-
ment effect in turn points towards constraints on the product market that do not allow for
sufficiently large labour demand responses to absorb the additional labour supply.
Because of the importance of the resident labour force’s skill composition as a source of
variation, I investigate whether initial relative skill shares have an independent effect on
future changes in labour market outcomes that could be driving the results but do not find
any indication for this. I also do not find evidence of any correlation between the popula-
tion growth rates of native Germans or foreign nationals and ethnic German immigration.
While the absence of a positive correlation can be seen as evidence for the effectiveness
of the allocation policy in preventing ethnic German immigrants to move to particularly
47
attractive labour markets, the absence of a negative correlation suggests that there is no
systematic out-migration of either native Germans or foreign nationals in response to the
immigrant inflows. This last result also holds when I look at skill-specific out-migration.
My estimates of the labour market impacts of immigration are therefore unlikely to be
underestimated as a result of unaccounted compensating migration flows.
Apart from estimating the short-run labour market effects of immigration in Germany,
this study also emphasises the importance of the existing structure of a labour market in
determining the effect of an immigrant inflow using spatial correlations. An identical rel-
ative inflow of immigrants into two regions will have substantially different impacts on
local labour market outcomes if these regions differ in terms of their existing skill mix.
In the context of a governmental allocation policy such as the one described in this paper,
an even distribution in terms of numbers of immigrants relative to the existing population
does therefore not necessarily lead to an even distribution of their labour market effects
across regions.
While this study has focussed on the impact of an exogenous inflow of immigrants on
relative labour market outcomes, an interesting avenue to pursue in the future could be to
look at changes in absolute terms. The arrival of new immigrants will typically lead to
a redistribution in an economy with a net positive effect on national income accruing to
the resident population, the immigrant surplus, as long as the immigrants differ from the
resident population in terms of their skills and lower their wages (Borjas, 1995). In theory,
the more different the immigrants are from the existing workforce, the larger should be
the immigrant surplus they give rise to in a region. The allocation policy described in
this paper offers a good framework for studying this theory. As opposed to cross-country
studies, the major advantage of the German context is that both the actual immigrant
inflows and the existing institutional settings are homogenous across regions, making it
easier to isolate the mechanism by which immigrant inflows lead to immigrant surpluses.
48
References
Altonji, J. G. and D. Card (1991). The Effects of Immigration on the Labor Market Out-
comes of Less-skilled Natives. In J. M. Abowd and R. B. Freeman (Eds.), Immigration,
Trade, and the Labor Market, Chapter 7, pp. 201–234. Chicago: University of Chicago
Press.
Angrist, J. D. and A. D. Kugler (2003). Protective or Counter-Productive? Labour Market
Institutions and the Effect of Immigration on EU Natives. Economic Journal 113(488),
F302–F331.
Bauer, T., B. Dietz, K. F. Zimmermann, and E. Zwintz (2005). German Migration: De-
velopment, Assimilation, and Labour Market Effects. In K. F. Zimmermann (Ed.),
European Migration, Chapter 7, pp. 197–261. Oxford: Oxford University Press.
Bauer, T. and K. F. Zimmermann (1997). Unemployment and Wages of Ethnic Germans.
Quarterly Review of Economics and Finance 37, 361–377.
Bauer, T. and K. F. Zimmermann (1999). Occupational Mobility of Ethnic Migrants. IZA
Discussion Paper No. 58.
Bender, S., A. Haas, and C. Klose (2000). The IAB Employment Subsample 1975-1995:
Opportunities for Analysis Provided by the Anonymised Subsample. Journal of Applied
Social Science Studies 120, 649–662.
Bonin, H. (2005). Wage and Employment Effects of Immigration to Germany: Evidence
from a Skill Group Approach. IZA Discussion Paper No. 1875.
Borjas, G. J. (1995). The Economic Benefits from Immigration. Journal of Economic
Perspectives 9(2), 3–22.
Borjas, G. J. (2003). The Labor Demand Curve is Downward Sloping: Reexamining the
Impact of Immigration on the Labor Market. Quarterly Journal of Economics 118(4),
1335–1374.
49
Borjas, G. J. (2006). Native Internal Migration and the Labor Market Impact of Immigra-
tion. Journal of Human Resources 41(2), 221–258.
Bundesagentur fur Arbeit (2004). Arbeitsmarkt 2003. Amtliche Nachrichten der Bunde-
sagentur fur Arbeit.
Butcher, K. F. and D. Card (1991). Immigration and Wages - Evidence from the 1980s.
American Economic Review 81(2), 292–296.
Card, D. (1990). The Impact of the Mariel Boatlift on the Miami Labor Market. Industrial
and Labor Relations Review 43(2), 245–257.
Card, D. (2001). Immigrant Inflows, Native Outflows, and the Local Labor Market Im-
pacts of Higher Immigration. Journal of Labor Economics 19(1), 22–64.
Card, D. and J. DiNardo (2000). Do Immigrant Inflows Lead to Native Outflows? Amer-
ican Economic Review 90(2), 360–367.
Carrington, W. J. and P. J. de Lima (1996). The Impact of 1970s Repatriates from Africa
on the Portuguese Labor Market. Industrial and Labor Relations Review 49(2), 330–
347.
Dustmann, C., F. Fabbri, and I. Preston (2005). The Impact of Immigration on the British
Labour Market. Economic Journal 115(507), F324–F341.
Dustmann, C. and A. Glitz (2005). Immigration, Jobs and Wages: Theory, Evidence and
Opinion. CEPR Report, London.
Edin, P.-A., P. Fredriksson, and O. Aslund (2003). Ethnic Enclaves and the Economic
Success of Immigrants - Evidence from a Natural Experiment. Quarterly Journal of
Economics 118(1), 329–357.
Friedberg, R. M. (2001). The Impact of Mass Migration on the Israeli Labor Market.
Quarterly Journal of Economics 116(4), 1373–1408.
50
Friedberg, R. M. and J. Hunt (1995). The Impact of Immigration on Host Country Wages,
Employment and Growth. Journal of Economic Perspectives 9(2), 23–44.
Gartner, H. (2004). Die Imputation von Lohnen oberhalb der Beitragsbemessungsgrenze
in der IAB-Beschaftigtenstatistik. IAB Working Paper.
Gaston, N. and D. Nelson (2002). The Employment and Wage Effects of Immigration:
Trade and Labour Economics Perspectives. In D. Greenaway, R. Upward, and K. Wake-
lin (Eds.), Trade, Investment, Migration and Labour Market Adjustment, Chapter 12,
pp. 201–235. Palgrave.
Gould, E. D., V. Lavy, and M. D. Paserman (2004). Immigrating to Opportunity: Estimat-
ing the Effect of School Quality Using a Natural Experiment on Ethiopians in Israel.
Quarterly Journal of Economics 119(2), 489–526.
Hartog, J. and A. Zorlu (2005). The Effect of Immigration on Wages in Three European
Countries. Journal of Population Economics 18(1), 113–151.
Hunt, J. (1992). The Impact of the 1962 Repatriates from Algeria on the French Labor
Market. Industrial and Labor Relations Review 45(3), 556–572.
Klose, H.-U. (1996). Bevolkerungsentwicklung und Einwanderungspolitik. In Integration
und Konflikt, Volume Nr. 69, pp. 14. Gesprachskreis Arbeit und Soziales der Friedrich-
Ebert-Stiftung.
LaLonde, R. J. and R. H. Topel (1991). Labor Market Adjustments to Increased Immi-
gration. In J. M. Abowd and R. B. Freeman (Eds.), Immigration, Trade, and the Labor
Market, Chapter 6, pp. 167–199. Chicago: University of Chicago Press.
Manacorda, M., A. Manning, and J. Wadsworth (2006). The Impact of Immigration on
the Structure of Male Wages: Theory and Evidence from Britain. CReAM Discussion
Paper No. 08/06.
OECD (2004). Employment Outlook 2004. OECD Publications, Paris.
51
Ottaviano, G. I. P. and G. Peri (2006). Rethinking the Effects of Immigration on Wages.
NBER Working Paper No. 12497.
Piil Damm, A. (2006). Ethnic Enclaves and Immigrant Labour Market Outcomes: Quasi-
Experimental Evidence. CReAM Discussion Paper No. 07/06.
Pischke, J.-S. and A. B. Krueger (1998). Observations and Conjectures on the U.S. Em-
ployment Miracle. In Third Public GAAC Symposium: Labor Markets in the USA and
Germany, pp. 99–126. Bonn: German-American Academic Council.
Pischke, J.-S. and J. Velling (1997). Employment Effects of Immigration to Germany: An
Analysis Based on Local Labor Markets. Review of Economics and Statistics 79(4),
594–604.
Salt, J. and H. Clout (1976). Migration in Post-War Europe: Geographical Essays. Lon-
don: Oxford University Press.
Welch, F. (1999). In Defense of Inequality. American Economic Review 89(2), 1–17.
Papers and Proceedings of the One Hundred Eleventh Annual Meeting of the American
Economic Association.
Winter-Ebmer, R. and J. Zweimuller (1996). Immigration and the Earnings of Young
Native Workers. Oxford Economic Papers 48(3), 473–491.
Winter-Ebmer, R. and J. Zweimuller (1999). Do Immigrants Displace Young Native
Workers: The Austrian Experience. Journal of Population Economics 12(2), 327–340.
52
Appendix A
Sample Description
All data on the local labour force is based on the IAB Employment Subsample 1975-2001.
This data set contains complete employment histories of 2% of all employees subject to
social security contributions in Germany, which translates into approximately 460,000 ob-
servations per year for West Germany. For each year, I collect the relevant information at
the cut-off date of 31 December. I delete all individuals that are marginally employed (ger-
ingfugig beschaftigt, pers gr=109, 209, 110, 202, 210) from the sample since these are
only recorded from 1999 onwards. I also delete observations that indicate a parallel em-
ployment spell (level26=0). I include only men and women aged 15 to 64. I impute missing
or unknown values for occupation, educational attainment and location of an individual
with the most recent information from previous spells of the same individual, if available.
Occupations are aggregated to five groups based on the American SF-3 Occupation Ta-
ble. The aggregation key can be obtained upon request. Education levels are aggregated to
three groups: “low” for individuals “without completed education” (bild=0), “without A-
levels and without vocational training” (bild=1), or “with A-levels but without vocational
training” (bild=3); “intermediate” for individuals “without A-levels but with vocational
training” (bild=2) or “with A-levels and with vocational training” (bild=4); and “high”
for individuals “with (technical) college degree” (bild=5, 6). Potential experience, which
is used in the regressions to obtain adjusted labour market outcomes, is calculated as cur-
rent year minus year of birth minus age at the end of educational/vocational training. The
average age for each education level is set at 15 for individuals “without completed ed-
ucation”, 16 for those “without A-levels and without vocational training”, 19 for those
“without A-levels but with vocational training” or “with A-levels but without vocational
training”, 22 for those “with A-levels and with vocational training”, and 25 for those
“with (technical) college degree” or unknown or missing values (which, based on their
average wage rate, seem most similar to college educated individuals). Foreign nationals
are aggregated to sixteen groups according to their countries or regions of citizenship:
53
Turkey, former Yugoslavia, Italy, Greece, Poland, the former Soviet Union, Portugal, Ro-
mania, Western Europe, Central & Eastern Europe, Africa, Central & South America,
North America, Asia, Australia & Oceania, and Others. Individuals are considered unem-
ployed if they are benefit receivers (typ1=6). For the construction of average wages I only
consider individuals that are working full-time (stib<5). All wages are converted into real
wages in Euros at constant 1995 prices using the German CPI for all private households.
Wage records that are right censored at the social security contribution ceiling are imputed
using a method developed by Gartner (2004). I aggregate the 326 West German counties
(excluding Berlin) to 204 labour market regions using an aggregation key provided by the
IAB.
Appendix B
Institutional Background
Table B-1: West Germany’s states and their implementation of the Assigned Place of Res-idence Act
No. oflabour State Actual Law Date of In In
No. of market quota quota imple- imple- unrestricted restrictedcounties regions in % 1996-2001 mented mentation sample sample
Schleswig-Holstein 15 7 3.3 3.4 yes 1.3.1996 yes yesHamburg 1 1 2.1 2.1 yes 1.3.1996 yes yesLower Saxony 46 35 9.2 8.2 yes 7.4.1997 yes yes, from 1997Bremen 2 0 0.9 0.9 yes 1.3.1996 yes yesNorth Rhine-Westphalia 54 36 21.8 21.6 yes 1.3.1996 yes yesHesse 26 16 7.2 7.2 yes 1.1.2002 yes noRhineland Palatinate 36 21 4.7 4.6 no - yes noBaden-Wurttemberg 44 29 12.3 12.1 yes 1.3.1996 yes yesBavaria 96 55 14.4 14.3 no - no noSaarland 6 4 1.4 1.4 yes 11.3.1996 yes yes
Overall 326 204 77.3 75.8 8/10 - 9/10 7/10
Notes: The labour market region in Hamburg also comprises three counties that are situated in Schleswig-Holsteinand one county that is situated in Lower Saxony. Because of the dominance of Hamburg’s and Schleswig-Holstein’scounties, this labour market region is already used from 1996 onwards when these two states adopted the AssignedPlace of Residence Act. There are two labour market regions in Lower Saxony that each comprise one of Bremen’scounties. Because each labour market region here consists of one county from Lower Saxony and one county fromBremen, I conservatively include these labour market regions only from 1997 onwards when Lower Saxony imple-mented the new legislation. Finally, there is one labour market region in Baden-Wurttemberg that comprises one ofBavaria’s counties. Because this labour market region consists of two counties from Baden-Wurttemberg and only onefrom Bavaria, I include it from 1996 onwards.
54
Appendix C
The Empirical Model
The empirical analysis in this paper is based on a theoretical model derived by Card (2001)
in which immigration impacts local labour markets by changing the relative supplies of
different skill groups. Suppose that a single output good Y is produced in labour market
region r in a given year t with a production function
Yrt = F(Krt ,Lrt),
where Krt are non-labour inputs and Lrt is a nested CES production function of different
skill groups j that are imperfect substitutes:
Lrt =(∑
j(e jrtN jrt)(σ−1)/σ
)σ/(σ−1).
Here N jrt is the number of individuals with skill level j employed in region r at time t and
σ is the elasticity of substitution between the different skill groups. e jrt reflect region-
and skill-specific productivity levels. If the wage rate of skill group j in region r at time t
is now given by w jrt and the selling price of output from region r in year t by qrt , equating
the marginal product of a skill group with its real product wage will lead to the following
expression:
logN jrt = θrt +(σ −1) loge jrt −σ logw jrt , (C-1)
where θrt = σ log [qrtFL(Krt ,Lrt)L1/σrt ] is a region- and time-specific component shared by
all skill groups. Let Pjrt be the labour force of individuals in skill group j in labour market
region r in year t and assume a log-linear labour supply function
log(N jrt/Pjrt) = ε logw jrt (C-2)
55
with ε > 0. Then using Equations C-1 and C-2, I can obtain the following expressions for
the employment/labour force and average wage rate of skill group j in region r at time t:
log(N jrt/Pjrt) = ε/(ε +σ){(θrt − logPrt)+(σ −1) loge jrt − log(Pjrt/Prt)},
logw jrt = 1/(ε +σ){(θrt − logPrt)+(σ −1) loge jrt − log(Pjrt/Prt)},
where Prt is the overall labour force in labour market region r at time t.36 Both local
wages and employment rates are determined by three factors: a common region- and
time-specific component, a skill-, region- and time-specific productivity component, and
the relative labour force shares of the different skill groups. If I decompose the unobserved
productivity component into four parts
loge jrt = e jr + e jt + ert + e′jrt ,
where e jr represents skill- and region-specific effects, e jt is a skill- and time-specific
effect, ert is a region- and time-specific effect, and e′jrt is a skill-, region- and time-specific
effect, I can obtain two regression models for the employment and wage rates:
log(N jrt/Pjrt) = v jr + v jt + vrt +β1 log f jrt + v jrt ,
logw jrt = u jr +u jt +urt +β2 log f jrt +u jrt ,
where f jrt = Pjrt/Prt denotes the fraction of the overall labour force in labour market r
at time t that falls into skill group j. Finally, taking first differences provides the set of
36I use the labour force rather than the working age population for Pjrt and Prt . I am therefore not able tocapture responses through entries to or exits from the labour force which, while less an issue for men, maybe problematic when looking at female labour market outcomes.
56
equations that are the basis of the empirical analysis in this paper:
∆ log(N jrt/Pjrt) = v′jt + v′rt +β1∆ log f jrt +∆v jrt ,
∆ logw jrt = u′jt +u′rt +β2∆ log f jrt +∆u jrt ,
where v′jt , u′jt , v′rt , and u′rt are interactions of skill and year fixed effects and region and
year fixed effects, respectively, and ∆v jrt and ∆u jrt are unobserved error components that
depend on the productivity terms e′jrt and e′jrt−1.
57
Appendix D
Tables
Table D-1: Impact of changes in relative factor shares on the employment/labour forcerate by gender
Men WomenAll regions Restricted regions All regions Restricted regions
OLS IV OLS IV OLS IV OLS IV
(1) (2) (3) (4) (5) (6) (7) (8)
Occupation groups
All -0.117*** -0.281 -0.121*** -0.369* -0.130*** 0.186 -0.121*** -0.283(.014) (.862) (.017) (.218) (.020) (.314) (.023) (.324)
[0.76] [2.90] [3.43] [2.94]
All unweighted -0.120*** 0.118 -0.130*** -0.370* -0.137*** 0.782 -0.109*** -0.072(.014) (.463) (.017) (.200) (.030) (1.711) (.025) (.388)
[1.38] [2.98] [1.33] [2.98]
All aged 25-54 -0.108*** -0.317 -0.119*** -0.305* -0.128*** 0.282 -0.116*** -0.315(.014) (.396) (.016) (.158) (.022) (.441) (.023) (.413)
[1.70] [3.69] [2.98] [2.53]
Germans only -0.127*** -0.083 -0.131*** -0.375* -0.118*** -0.129 -0.109*** -0.169(.014) (.400) (.017) (.206) (.021) (.322) (.022) (.298)
[1.41] [3.29] [3.31] [3.56]
Observations 4440 4440 3185 3185 4436 4439 3185 3185
Education groups
All -0.035** -0.262** -0.043* -0.395 -0.130*** -0.283 -0.137*** -0.630(.014) (.119) (.022) (.251) (.039) (.737) (.036) (.652)
[4.61] [2.24] [1.78] [2.31]
All unweighted -0.030** -0.233 -0.036* -0.224 -0.158*** -0.750 -0.114*** -0.241(.012) (.163) (.019) (.144) (.051) (.605) (.036) (.292)
[2.95] [3.21] [2.52] [3.21]
All aged 25-54 -0.026* -0.135* -0.026 -0.405 -0.126*** -0.274 -0.141*** -0.601(.014) (.075) (.021) (.262) (.046) (.740) (.037) (.554)
[4.77] [2.15] [1.86] [2.75]
Germans only -0.039*** -0.209 -0.036* -0.224 -0.161*** 0.429 -0.114*** -0.241(.015) (.136) (.019) (.144) (.048) (1.547) (.036) (.292)
[3.99] [3.21] [1.33] [3.21]
Observations 2664 2664 1911 1911 2660 2660 1911 1911
Notes: see Table 5.
58
Table D-2: Impact of changes in relative factor shares on log daily wages by gender
Men WomenAll regions Restricted regions All regions Restricted regions
OLS IV OLS IV OLS IV OLS IV
(1) (2) (3) (4) (5) (6) (7) (8)
Occupation groups
All -0.041*** -0.754 -0.042** 0.100 -0.096** -0.165 -0.153*** -0.883(.015) (1.216) (.017) (.265) (.037) (.649) (.038) (.652)
[0.73] [2.11] [3.09] [2.85]
All unweighted -0.029* -0.235 -0.032** 0.045 -0.108** -0.776 -0.164*** -1.289*(.016) (.424) (.016) (.217) (.049) (3.055) (.051) (.767)
[1.38] [2.98] [1.43] [2.98]
All aged 25-54 -0.031** -1.033 -0.029* -0.231 -0.134*** -0.837 -0.188*** -0.968(.015) (1.054) (.016) (.222) (.041) (.655) (.041) (.707)
[1.21] [2.78] [2.71] [2.62]
Germans only -0.029* -0.445 -0.029* -0.024 -0.091** -0.595 -0.145*** -1.365*(.015) (.545) (.017) (.233) (.039) (.907) (.041) (.729)
[1.25] [2.51] [2.49] [2.92]
Observations 4440 4440 3185 3185 4431 4431 3185 3185
Education groups
All -0.031 0.026 -0.040 0.137 -0.065 1.585 -0.047 0.539(.030) (.092) (.027) (.278) (.046) (1.273) (.060) (.662)
[4.94] [2.35] [2.02] [2.13]
All unweighted -0.032 0.064 -0.055** 0.032 -0.087* 1.704** -0.076 0.368(.024) (.133) (.026) (.150) (.048) (.803) (.055) (.327)
[2.95] [3.21] [2.57] [3.23]
All aged 25-54 -0.018 0.007 -0.028 0.033 -0.035 0.010 -0.075 0.267(.033) (.109) (.027) (.308) (.056) (.892) (.072) (.789)
[4.26] [2.29] [1.51] [2.14]
Germans only -0.035 0.066 -0.055** 0.032 -0.061 1.318** -0.076 0.368(.029) (.101) (.026) (.150) (.049) (.639) (.055) (.327)
[4.75] [3.21] [2.77] [3.23]
Observations 2664 2664 1911 1911 2646 2653 1907 1907
Notes: see Table 6.
59
Table D-3: Impact of initial skill shares on labour market outcomesby gender, 1985 to 1987
∆(N jrt/Pjrt) ∆ logw jrt
Independent variable Occupation Education Occupation Education
Men
Initial skill share 0.017 -0.014 0.017 0.010(.015) (.021) (.020) (.020)
Observations 1480 888 1480 888R2 0.68 0.64 0.72 0.84
Women
Initial skill share 0.021 0.008 -0.082** -0.100(.024) (.023) (.039) (.061)
Observations 1478 886 1476 876R2 0.86 0.68 0.68 0.99
Notes: see Table 7.
60